December 2019 Authors David Sandalow Center on Global Energy Policy, Columbia University Chair, ICEF Innova on Roadmap Project Julio Friedmann Center on Global Energy Policy, Columbia University Roger Aines Lawrence Livermore Na onal Laboratory Colin McCormick Walsh School of Foreign Service, Georgetown University Sean McCoy Department of Chemical and Petroleum Engineering, University of Calgary Joshuah Stolaroff Lawrence Livermore Na onal Laboratory Roger Aines and Joshuah Stolaroff contributed to the technical evalua ons but not the policy recommenda ons in this document.

This roadmap was prepared to facilitate dialogue at the Sixth Innova on for Cool Earth Forum (Tokyo October 2019), for fi nal release at COP-25 (Madrid December 2019). We are deeply grateful to the Ministry of Economy, Trade and Industry (METI) and New Energy and Industrial Technology Development Organiza on (NEDO), Japan, for launching and suppor ng the ICEF Innova on Roadmap Project of which this is a part. Contents

EXECUTIVE SUMMARY

CHAPTER 1 1 INTRODUCTION

CHAPTER 2 5 TECHNOLOGY OPTIONS FOR LOW-CARBON INDUSTRIAL HEAT 5 Hydrogen 11 Biomass 16 Electrifi cation 21 Carbon Capture, Use and Storage

CHAPTER 3 25 SECTORAL STUDIES 25 Cement 34 Iron and Steel 43 Chemicals

CHAPTER 4 53 INNOVATION PATHWAYS

CHAPTER 5 60 POLICY

CHAPTER 6 66 FINDINGS AND RECOMMENDATIONS

December 2019 iii “1.0” analysis of the topic. A er providing general EXECUTIVE background, we discuss four technological approaches for providing low-carbon industrial heat: hydrogen, biomass, electrifi ca on and CCUS. We next examine SUMMARY decarbonizing heat produc on in the cement, iron and steel, and chemical industries. We then turn to policy op ons and an innova on agenda. We close with INTRODUCTION fi ndings and recommenda ons. Roughly 10% of global greenhouse gas (GHG) emissions come from the produc on of heat for industrial processes—more than cars and planes TECHNOLOGY combined. Decarbonizing industrial heat produc on will be essen al to mee ng the Paris Agreement OPTIONS FOR LOW- goals. This topic has received far less a en on than decarboniza on of the power, transport or building CARBON INDUSTRIAL sectors. HEAT Most heavy industries require enormous quan es of heat. In many cases (including the cement, iron and Hydrogen. steel, and chemical industries), core processes involve Hydrogen combus on produces heat without carbon smel ng ore, breaking strong chemical bonds and/or dioxide (CO ) emissions. Hydrogen can subs tute for increasing the energy content of products. 2 coal, oil and gas as a fuel in some industrial processes, Today, almost all industrial heat is provided by reducing on-site emissions. However the produc on combus on of coal, oil or natural gas. These fossil fuels of hydrogen may involve substan al CO2 emissions, provide the high temperatures, con nuous opera on reducing or elimina ng the CO2 benefi ts of switching to and reliability many industrial processes require. Any hydrogen. Low-carbon produc on of hydrogen is thus op ons for decarbonizing industrial heat must match essen al for hydrogen to play a role in decarbonizing these capabili es or be part of a broader change in industrial heat. industrial processes. The most common hydrogen produc on process today Op ons to provide low-carbon heat for industry include is steam methane reforming (“gray” hydrogen), which hydrogen; biomass; electrifi ca on; carbon capture, use has signifi cant CO2 emissions. This can be substan ally and storage (CCUS); nuclear power; and concentrated decarbonized by adding CCUS (“blue” hydrogen), solar power (CSP). Few if any of these op ons are well reducing the carbon footprint by 55-90% or more. Low- developed in the context of industrial heat produc on. carbon hydrogen can also be made through electrolysis Several characteris cs of heavy industries create using zero-carbon power (“green” hydrogen). challenges in decarbonizing industrial heat produc on. Hydrogen could be used in many exis ng industrial First, industrial facili es are long-lived capital stock, hea ng systems with small changes, especially for las ng decades. Second, many industrial products are chemical synthesis. Issues such as sensors, controls, globally traded commodi es, subject to signifi cant loss corrosion and embri lement appear resolvable with of market share due to small increases in produc on minor costs and system modifi ca ons. Blue hydrogen costs. Third, many industrial facili es are far from would add modest costs to produc on of hydrogen renewable resources such as biomass or abundant solar and raw industrial products (20-50% increase). Green radia on, limi ng decarboniza on op ons. Fourth, many hydrogen would add substan al costs (200-400% governments view these industries as core na onal increase). As costs for fi rm renewable power decrease in assets, aff ec ng na onal security and the balance of the future, green hydrogen may become more a rac ve trade. and could take advantage of infrastructure originally This Roadmap explores the challenge of industrial installed to use blue hydrogen. heat decarboniza on. It is intended to be an ini al,

iv Biomass. grid infrastructure upgrades are needed for large-scale Biomass provides considerable heat when burned. industrial electrifi ca on. Biomass can be converted to useful intermediates such as biomethane, biodiesel and bio-char, and provides Carbon capture, use and storage (CCUS). a carbon source and chemical reductant important in CCUS has an important role to play in reducing emissions some industries. Biomass has the poten al to deliver net from produc on of industrial heat. The building blocks

low-carbon heat, since biomass can regrow, absorbing of CCUS include separa on of CO2 from combus on

CO2 released during combus on. However land use products or hydrocarbon fuels, transporta on of CO2 changes related to biomass harves ng can reduce or to a suitable storage site (or loca on where it is used),

eliminate these CO2 benefi ts. Transport and processing and geologic storage of CO2 or conversion of CO2 into a of biomass, as well as use of fer lizer, can also reduce range of products (e.g., carbonate minerals, chemicals the GHG benefi ts of biomass combus on. and fuels). CCUS is a rac ve because it usually does not require wholesale changes to the underlying industrial Approximately 200-500 EJ/y of sustainably produced processes. biomass can be available by 2050, similar to the projected global industrial energy demand of 330 EJ/y Experience with CCUS has grown considerably since in 2040. Nevertheless, scaling biomass suffi ciently to 1996, when the fi rst “purpose built” project began

play a signifi cant role in industrial heat produc on storing CO2 captured from a gas processing pla orm would be a challenge. Biomass is more geographically deep under the North Sea. Today, CCUS projects

diverse and expensive to collect and transport than are capturing CO2 that would have otherwise been fossil fuels. Woody biomass has about half the energy emi ed from power genera on, ethanol fermenta on, density and considerably lower bulk density (before gas separa on, iron and steelmaking, and hydrogen

grinding) than coal. There are compe ng demands for produc on. CO2 capture for industrial processes— biomass in a low-carbon future, including as vehicle par cularly cement and steel—requires further fuel, dispatchable electricity and means of nega ve development through demonstra on projects at scale. emissions. Despite these challenges, biomass has the Infrastructure is needed to transport and geologically

poten al to contribute to low-carbon heat for industry in store large volumes of CO2. some applica ons.

Electrification. SECTORAL STUDIES A wide variety of exis ng and emerging electrical Cement. technologies can provide high-temperature industrial process heat, including resistance hea ng, microwaves, Cement provides the founda on for the built induc on and electric arc furnaces. Electrical hea ng has environment. Currently, over 4 Gt of cement are high controllability of temperature and dura on of heat produced annually, resul ng in more than 2 Gt per applica on, rela vely low maintenance, and inherently year of CO2 emissions. CO2 emissions from cement low emissions when powered by low-carbon electricity. manufacturing result not only from high-temperature However, reliable electricity in industrially relevant heat—nearly 1,500 °C in the cement kiln—but also from decomposi on of limestone (CaCO ). Many strategies quan es is not always available and in general is higher 3 cost than combus on-based technologies. for reducing these emissions have been considered, including fuel switching in conven onal cement making, The installa on of electric process heat systems o en fundamental changes in the composi on of cement and requires more changes to exis ng equipment than more effi cient use of concrete in design. switching to alternate combus on-based fuels (such as hydrogen or biomass). It may also require substan al Subs tu on of lower-carbon-intensity fuels for coal plant redesign. The use of electricity in industrial is already having a substan al impact in the cement process heat applica ons can place major burdens sector. This could be furthered by increased use on the electric grid. While some op miza on such of biomass-based wastes and sustainable biofuels. as par cipa on in demand-side management (DSM) However, given the limited supply of sustainable biomass systems is possible, this is limited in prac ce and major and compe on that may emerge for its diff erent

v uses, this may not be cost-eff ec ve in large quan es. Hydrogen (blue or green) appears to be the most ready CCUS appears to be an important op on for reducing subs tute for current fossil fuel heat sources, in large emissions from cement produc on. part because many chemical facili es use predominantly gaseous fuels for heat produc on. CCUS retrofi ts may Iron and steel. cost less than switching to hydrogen in some instances,

The iron and steel sector is one of the largest in the although capturing CO2 from many distributed heat world, responsible for 7-9% of global direct emissions sources may prove challenging. Effi ciency improvements from fossil fuel. New steel produc on primarily uses a provide near-term opportuni es to reduce emissions. blast furnace to convert iron ore to pig iron, followed In the future, electrifi ca on may prove workable, by a basic oxygen furnace to convert pig iron to steel. especially for straigh orward subs tu ons such as This process is emissions-intensive, with most emissions steam produc on. To achieve widespread electrifi ca on, coming from the blast furnace. Electric arc furnaces low-carbon electric power would need to be much convert recycled steel and iron from other processes to cheaper and more readily available, and novel systems liquid steel in a far less emissions-intensive manner but of heat deposi on would require development and are limited by availability of recycled material. demonstra on. In blast furnaces, process heat is provided by combus on of coke. Coke also provides carbon as a INNOVATION reductant, acts as structural support to hold the ore The greatest challenge in innova on for industrial burden, and provides porosity for rising hot gas and heat is the extreme diversity of processes that require sinking molten iron. Because of these mul ple roles, carbon-free energy. Systema c changes across the directly replacing coke combus on with an alterna ve economy, such as switching from methane to hydrogen source of process heat is not prac cal. Op ons for in gas pipelines, will be important op ons, but there reducing process-heat-related emissions from blast is currently no analy cal basis on which to compare furnaces include reducing coke through hydrogen op ons. Detailed and cross-cu ng analysis of the co-injec on and plasma torch super-hea ng of hot blast benefi ts and costs of zero-carbon fuels, biomass, and air. Direct-reduced iron (DRI) and smel ng reduc on hybrid systems involving CCUS and direct air capture iron can also be used in blast or arc furnaces to reduce are needed before na ons can commit to wholesale emissions. Biomass-nugget smel ng, hydrogen- industrial changes. reduc on iron making, and electroly c steel produc on also off er alterna ves. CCUS can signifi cantly reduce The most immediate pathway to decarbonizing industrial process-heat-related emissions when applied to fl ue gas heat is likely to be one that incrementally reduces at an integrated steel mill, blast furnace gas, or DRI and emissions, with rela vely small changes from op ons smel ng reduc on processes. like more effi cient heat applica on, reduced carbon footprint of fuels, and hybrids involving par al carbon Chemicals. capture and storage (CCS). It is important to evaluate The global chemical industry is roughly 3% of global and develop these par al pathways in concert with those that completely transform industrial processes. CO2 emissions. Energy demand for chemicals is greater than for either cement or steel, refl ec ng enormous Zero-carbon fuels are likely to be an important element heat consump on. Chemical facili es use a wide set of of such transi ons. Global transport of hydrogen and feedstocks and fuels, consuming natural gas, natural gas biomass needs to be evaluated for both economic and liquids, heavy hydrocarbons (e.g., bitumen, asphalt), coal climate impacts to determine if that is an approach that and hydrogen. Facili es are highly complex with a wide can overcome regional shortages in those two fuels. The range of chemical-produc on processes. Many reac ons safety and opera onal issues of zero-carbon fuels (fl ame require fi t-for-purpose reactors that cannot readily be visibility for hydrogen, methane leakage for renewable replaced. A concern specifi c to the chemical industry is natural gas and food/ecosystem tradeoff s for biofuels) the broad distribu on of heat sources, which can include need to be carefully evaluated. dozens or hundreds of small emissions sources such as High capital costs are likely to be a major barrier to the burners, furnaces and boilers. transi on to zero-carbon industrial heat sources. Finally,

vi the costs of completely decarbonizing by u lizing direct 6. Carbon tariff s. Carbon tariff s (some mes called air capture need to be evaluated for industries that have “carbon border-tax adjustments”) are a tool for achieved par al decarboniza on but face massive capital addressing interna onal compe veness concerns. A expenditures to completely eliminate their carbon country that requires its manufacturers to transi on emissions. to low-carbon industrial heat could tax imports of relevant products from countries that fail to do so. 7. Mandates. Governments could prohibit the use POLICY of fossil fuels or require the use of low-carbon Policy tools are essen al for decarbonizing industrial technologies for genera ng heat in certain industrial heat, both in the short- and long-term. sectors. 8. Voluntary industry associa ons. Industry Market forces alone are insuffi cient, since CO emi ers 2 associa ons such as World Steel Associa on, World do not bear the full costs of their emissions. Government Petroleum Council, World Cement Associa on and policies are essen al. Many policy tools are available to World Business Council on Sustainable Development help with decarbonizing industrial heat. These include: can help develop methods and standards for 1. Government support for research and decarbonizing industrial heat. They can play an development (R&D). Na onal governments spend important role in informa on-sharing on such topics. roughly $15 billion annually on R&D for clean energy 9. Clean Energy Ministerial. The Clean Energy technologies. These programs have played important Ministerial is a global forum where major economies roles in the development of countless technologies work together to share best prac ces and promote in recent decades. Increased R&D funding on policies and programs that encourage and facilitate industrial heat decarboniza on is essen al. the transi on to a global clean energy economy. A 2. Government procurement. Governments are major Clean Energy Ministerial ini a ve on industrial heat purchasers of steel, cement, chemicals and other decarboniza on could help to share best prac ces products that require heat in the manufacturing and accelerate their adop on. process. Procurement standards that give preferences to products with the lowest embedded carbon content could drive signifi cant changes in FINDINGS AND industrial behavior. 3. Fiscal subsidies. Decarbonizing industrial heat will RECOMMENDATIONS impose costs on aff ected businesses. Government Decarbonizing industrial heat produc on will require policies can help to reduce those costs with fi scal innova ng in mul ple sectors. Our key fi ndings and subsidies. These can take several forms, including tax recommenda ons include: incen ves, grants, loan guarantees, feed-in-tariff s and contracts for diff erences. ■ Finding 1: Emissions from industrial heat produc on 4. Infrastructure development. The transi on limit progress on climate goals. to low carbon industrial heat may require new ■ Finding 2: The opera onal requirements and infrastructure (such as electric transmission lines commercial reali es of many industries limit or hydrogen pipelines). Governments can play a opportuni es for decarboniza on. central role in facilita ng the development of such ■ Finding 3: There are few op ons today for low-carbon infrastructure through permi ng, fi nancing and heat genera on for industry. other measures. ■ Finding 4: Exis ng op ons face challenges based on price, performance and viability. 5. Carbon prices. A price on CO2 emissions, whether through an emissions-trading program or tax ■ Finding 5: There appear to be many pathways to mechanism, provides emi ers with an important improving cost, performance and viability of low- incen ve to cut emissions. The carbon prices that carbon industrial heat op ons. might be needed to induce a transi on from fossil ■ Finding 6: Many policy op ons exist that could fuels for industrial heat produc on are unclear. improve the speed and magnitude of industrial

vii decarboniza on and deployment of low-carbon alterna ve heat systems. ■ Recommendation 1: Key stakeholders should priori ze industrial heat produc on as a key element of any climate mi ga on strategy. ■ Recommendation 2: Industry-specifi c analy cal frameworks and innova on agendas are essen al. Governments and companies together should develop new ini a ves and R&D programs to focus on industrial sector decarboniza on with a focus on heat supplies. ■ Recommendation 3: Governments should iden fy and implement a set of policy ac ons to accelerate and support industrial decarboniza on, star ng with “buy clean” procurement. Final thoughts This Roadmap is an ini al foray into this extremely important and complex topic. One core fi nding of this Roadmap is that more work is needed on this topic. The urgency of climate change requires profound and rapid ac on. More data, input and technology op ons for decarbonizing industrial heat are urgently needed.

viii demand reduc on and development of a circular CHAPTER 1 economy, the ambi ous goals of the Paris Agreement will be diffi cult or impossible to achieve without signifi cant reduc ons in emissions from industrial heat INTRODUCTION produc on. Roughly 10% of global greenhouse gas (GHG) emissions Requirements come from the produc on of heat for industrial Most heavy industries require enormous quan es of processes. This is more than GHG emissions from cars heat at high temperature. In many cases (including the and planes combined.1,2 Decarbonizing industrial heat cement, iron and steel, and chemical industries), the produc on will be essen al to mee ng the goals set core industrial processes involve smel ng ore, breaking forth in the Paris Agreement, including achieving net strong chemical bonds or increasing a product’s energy zero emissions in the second half of this century, yet content. These processes produce substan al GHG technological approaches for decarbonizing industrial emissions. Earlier this century, industrial emissions heat produc on are far from maturity. This topic has growth was driven by high temperature in sectors received far less a en on than decarboniza on of the including cement and steel (in part to build Chinese power, transport and building sectors. megaci es). Although these sectors s ll produce large This Roadmap explores the challenge of industrial heat emissions, current growth is from medium temperature decarboniza on. A er providing background, we explore applica ons including refi ning and chemicals. four technological approaches with the poten al for The needs of the specifi c industries themselves genera ng industrial heat without GHG emissions: vary considerably and are extremely heterogenous hydrogen; biomass; electrifi ca on; and carbon capture, (even within one major produc on facility). Three use and storage. We next explore the poten al for requirements are key: decarbonizing heat produc on in three industries: cement, iron and steel, and chemicals. We then turn to ■ Temperature: Industrial products are made through policy op ons and an innova on agenda. We close with the applica on of high-grade heat to feedstocks. fi ndings and recommenda ons. Temperature demands vary signifi cantly from around 200 °C to nearly 2,000 °C (Table 1.1). Background ■ Flux: Industrial heat demands must be met with high Almost a quarter of global GHG emissions come from (and commonly con nuous) heat fl ux into the system. the industrial sector.3,4 The cement, steel and chemical The fl ux must be large enough to sustain reasonable industries are the largest contributors. The refi ning, produc on (Table 1.1). fer lizer and glass industries are signifi cant contributors ■ Reliability: Most heavy industrial produc on occurs as well. In 2018, global industrial GHG emissions grew at at large facili es with high capital costs (e.g., refi nery, a rate of 2.7%.5 steel mill). Most of these facili es operate with very high capacity factors, commonly 60-95%.7 As such, Recent studies have provided excellent analyses of heat supply must be dispatchable and available both several strategies for reducing emissions, including throughout the day and throughout the year. process intensifi ca on, material subs tu on, overall demand reduc on and energy effi ciency.6 Rela vely li le Any viable op on to replace exis ng sources of industrial a en on has focused on how industry uses heat. heat must be able to achieve the temperatures, In many industrial applica ons, high quality heat is the fl uxes and reliability necessary for robust, con nuous most important input a er primary feedstocks. Heat opera on. Approaches or fuels that cannot achieve high produc on, usually through fossil fuel combus on, is temperatures (e.g., heat pumps) or are intermi ent either the largest contributor or the second largest (e.g., hea ng with variable renewable power) will have contributor to industrial sector emissions. Together, limited u lity as viable subs tutes. Finally, most op ons industrial heat represents roughly 40% of total industrial must operate in the specifi c geography where these emissions. Even with substan al effi ciency gains, industries exist and operate (e.g., along the Gulf of Mexico or Northern European ports).

December 2019 1 Temperature Energy requirements Applica on & process required (°C) (GJ/ton product) Glass: Silica mel ng 1,600 ≈3 Cement: Clinker produc on 1,450 4 Steel: Blast furnace opera on 1,100 11-14 Hydrogen produc on: Steam methane reforming 820 16 Fer lizer: Ammonia synthesis 450 36 Petrochem: Methanol synthesis 300 33

Table 1.1. Temperature requirements and energy requirements per ton produc on for key industrial processes & applica ons (global averages)

Options of industrial applica ons based on temperature Op ons for providing low-carbon heat of industrial requirements alone (Figure 1.1). quality are not well developed. Challenges can be Some low-carbon op ons will only prove viable for physical or chemical (e.g., the temperature of steam at a small number of industrial applica ons. Stated conven onal nuclear power plants), geographic (e.g., diff erently, it is unlikely that one pathway will prove the availability of solar radiance or biomass feedstock) superior to the others in all contexts. A por olio of and economic (e.g., the cost of electroly c produc on of fuels is likely to serve a range of poten al industrial hydrogen). applica ons, and individual sectors or plants may select Unfortunately, no op on exists that can be widely diff erent op ons based on their geography, policy deployed today. Current op ons include: framework and asset requirements. ■ Hydrogen combustion: Burning hydrogen made from Nature of industrial operations and markets electrolysis or decarbonized fossil fuels. Several characteris cs of industrial opera ons and ■ Biomass combustion: Burning of unrefi ned biomass markets create challenges in decarbonizing industrial (e.g., agricultural wastes and wood pellets). heat produc on.8 ■ Biofuel combustion: Burning of refi ned biomass (e.g., biogas, biodiesel and corn ethanol). First, industrial facili es are long-lived capital stock. ■ Electrical heating: Direct and indirect hea ng Turnover o en takes place over many decades. approaches such as resis ve hea ng, induc on Components such as rotary cement kilns, blast furnaces, hea ng and dielectric hea ng (e.g., microwaves). cataly c crackers and hydrogen produc on units can individually cost $100s of millions and are central to the ■ Carbon capture, use and storage: Capturing CO opera on of mul -billion dollar assets. It may take 30-60 from combus on of fossil fuels (or biofuels) and years to replace core components of a large industrial sequestering it underground or in durable products. facility, and some facili es have operated for over 80 ■ Concentrated solar power: Large facili es (e.g., years and are s ll making products and revenues. This power towers) and more distributed approaches (e.g., creates a high hurdle rate to rebuilding and replacing small parabolic mirrors). high-emi ng units and aff ects the ability of innova ve ■ Conventional and advanced nuclear heat: Light solu ons to propagate into the sector. water reactors, emerging small modular reactors and advanced nuclear processes (e.g., sodium-cooled fast Second, many industrial products are globally traded reactors or nuclear fusion). commodi es (unlike electricity, which serves local or a In considering these op ons, heat quality, cost, regional markets). This means that prices of many availability and carbon footprint are all important. industrial products are set by interna onal trade. Small Some of these op ons can serve only a limited number a Cement and concrete markets are excep ons, since they are mostly used locally (although that has begun to change).

December 2019 2 Figure 1.1. Temperature requirements of key industrial process and the temperature limits provided by some op ons for low-carbon heat source replacements.8

increases in produc on costs could lead to a drama c Strategies for reducing GHG emissions from industrial loss of market share and loss of compe veness overall, heat produc on will be more likely to succeed if they which can aff ect the na onal trade balance and the take account of these quali es of the key sectors. overall health of the industry. This has led to narrow margins9,10 and a reluctance to increase costs. Framing for this Roadmap Third, many industrial facili es are located far from More and be er op ons for low-carbon industrial heat renewable resources such biomass or solar radia on are essen al to mee ng global climate goals. To improve suffi cient for concentrated solar power. O en, heavy the exis ng op ons and develop addi onal pathways industrial manufacturing facili es are found in ports to requires informa on and knowledge that is unavailable. facilitate trade and delivery of feedstocks. For example, The general lack of knowledge and informa on around over 70% of the US refi ning capacity lies along the Gulf the topic of industrial decarboniza on (and industrial of Mexico coast in Texas and Louisiana. heat in par cular) prevents investors, operators and policy makers from considering and implemen ng Fourth, many governments consider these industries alterna ves. This Roadmap explores op ons available to be core na onal assets, aff ec ng na onal security today as an early foray into the subject. and the balance of trade. In part for that reason, these industries have some mes received exemp ons This Roadmap is intended to be an ini al, “1.0” analysis or waivers from carbon pricing and environmental of op ons for decarbonizing industrial heat. In preparing regula ons. In some cases, this has been a factor this Roadmap, we have priori zed several technological contribu ng to signifi cant overcapacity (e.g., in Chinese pathways (hydrogen, biomass, electrifi ca on and CCUS). and Korean steel produc on). While some of this We do not explore other op ons including concentrated overcapacity may lead to closures, it may also lead to life solar power, gene cally modifi ed organisms and nuclear extension for lowest cost assets, which may have a poor heat. Our selec on was based on factors including carbon emissions profi le. geographic availability, technological readiness and

December 2019 3 public acceptance. We explore several key industries 4 IEA, 2018, “World Energy Outlook 2018”, h ps://www. (cement, iron and steel, and chemical) but have not iea.org/weo2018/ had the chance to explore others (including pulp and 5 R. Jackson et al., 2018, “Global energy growth is paper, glass making, and aluminum smel ng). We outpacing decarboniza on”, Environ. Res. Le ., h ps:// recognize that all viable op ons today are challenging, iopscience.iop.org/ar cle/10.1088/1748-9326/aaf303/ including most of those we have selected for analysis. meta The challenges include cost, availability, life-cycle carbon 6 Ins tute for European Studies, 2019, “Industrial footprint and engineering viability. Transforma on 2050: Towards and Industrial Strategy for a Climate Neutral Europe”, h ps://europeanclimate. Strategies for decarbonizing the industrial sector must org/wp-content/uploads/2019/04/Towards-an- also include effi ciency improvements, material use Industrial-Strategy-FULL-REPORT.pdf reduc on and development of a circular economy.6 7 Material Economics, 2019, “Industrial Transforma on Decarbonizing produc on of industrial heat will be an 2050: Pathways to net-zero emissions for European important part of the solu on set. Heavy Industy”, h ps://europeanclimate. org/wp-content/uploads/2019/04/Industrial- 1 M. Fischedick et al.,Climate Change 2014: Mi ga on Transforma on-2050.pdf of Climate Change (Fi h Assessment Report), Chapter 8 See J. Friedmann et al., 2019, “Low-carbon heat 10: Industry, Intergovernmental Panel on Climate solu ons for heavy industry: sources, op ons & costs Change, at p. 752, h ps://www.ipcc.ch/site/assets/ today”, Columbia Univ. Center on Global Energy Policy, uploads/2018/02/ipcc_wg3_ar5_chapter10.pdf h ps://energypolicy.columbia.edu/research/report/ 2 IEA, Tracking Clean Energy Progress, Transport Page, low-carbon-heat-solu ons-heavy-industry-sources- “Transport sector CO emissions by mode,” h ps:// op ons-and-costs-today www.iea.org/tcep/transport/ (accessed September 29, 9 Macrotrends, 2019a, U.S. Steel Profi t Margin 2006- 2019) 2019, h ps://www.macrotrends.net/stocks/charts/X/ 3 EPA, 2019, “Global Greenhouse Gas Emission Data”, united-states-steel/profi t-margins h ps://www.epa.gov/ghgemissions/global-greenhouse- 10 Macrotrends, 2019b, U.S. Concrete Profi t Margin 2006- gas-emissions-data 2019, h ps://www.macrotrends.net/stocks/charts/ USCR/u-s-concrete/profi t-margins

December 2019 4 combus on is 2800 °C.) Although today it is rare for CHAPTER 2 hydrogen combus on to create an industrial heat source, some applica ons burn hydrogen in boilers, stoves and vehicle engines. Hydrogen combus on TECHNOLOGY systems require special burners and in some cases require conversion from liquifi ed hydrogen to gas form. OPTIONS FOR LOW- In most other respects, hydrogen combus on for heat is extremely similar to burning natural gas and is a viable CARBON INDUSTRIAL subs tute for other gaseous fuels. Roughly half of hydrogen produced today worldwide is HEAT from natural-gas reforming.1 It is a mature technology and rela vely energy effi cient (65-75% conversion HYDROGEN effi ciency) and can operate wherever there is a natural Hydrogen is the most abundant element in the universe gas supply. Gas reforming itself uses high-temperature and extremely abundant on Earth. When burned, heat (700-1,000 °C) at elevated pressures (15-25 bars), hydrogen produces high-grade heat without carbon usually provided from natural gas furnaces. The reac on dioxide (CO2) emissions. Subs tu ng hydrogen for occurs in the presence of a catalyst. The fundamental hydrocarbon fuels such as natural gas is one poten al chemistry of reforma on can be represented in simple pathway for decarbonizing industrial heat produc on. terms: Although hydrogen is extremely abundant, it is usually Steam-methane reforming reaction bound to other elements in compounds such methane CH4 + H2O (+ heat) → CO + 3H2 (CH ) and water (H O). Separa ng hydrogen from these 4 2 Water-gas shift reaction compounds requires substan al amounts of energy CO + H2O → CO2 + H2 (+ small amount of heat) to break chemical bonds. The processes for doing so (chemical, electrical, thermal) are readily available today Although the fundamental chemistry is straigh orward, and used commercially in many industries in which the engineering is more complicated. The key process hydrogen is a feedstock. Pipelines in many countries (reforming) is strongly endothermic and consumes heat currently provide hydrogen as a feedstock to chemical (i.e., 206 kJ/mol). It operates at high temperature which and refi ning plants, steel plants and other industrial requires high combus on heat. Feedstock coming from facili es. a conven onal natural gas pipeline must be purifi ed, which requires separa ons of sulfur, nitrogen and other Hydrogen can be burned in air, producing a 2100 °C trace gases. These each have their own heaters, o en fl ame. (If burned in oxygen, the heat of hydrogen provided by pre-heaters and o en from heat recovery

Figure 2A-1. Hydrogen produc on processes. A: Steam methane reforming (SMR). B: Electrolysis of water.

December 2019 5 units. The fi nal separa ons through the pressure swing CO2 emissions from a facility (the rest comes from gas adsorp on unit also require heat and work. combus on in the hea ng systems). In addi on to steam-methane reforming (SMR), other Electrolysis of water is a completely diff erent op on for approaches include autothermal reforming (ATR), par al hydrogen produc on, using electricity as the energy oxida on and other more exo c methods (see below). source to break the chemical bonds in water, forming In refi neries, oil residues are commonly gasifi ed and hydrogen and oxygen (Figure 2A-1). The process refi nery gas streams are reformed. In loca ons where requires an electroly c cell and fairly pure water gas is expensive or supplies are limited (e.g., China, supplies. Electrolysis is typically more expensive than India, South Africa), coal or petcoke is gasifi ed as an gas reforming, with costs principally determined by the alterna ve feedstock to gas reforming and is commonly costs of electricity and electrolyzers. The electricity for combined with a water-gas shi reac on to maximize electrolysis can come from high-carbon or low-carbon produc on.a The dominant costs are the costs of gas sources. Electrolysis itself produces no greenhouse gases (85% of total) and the heavy industrial equipment used (GHGs). in reforma on. Importantly, CO is a direct chemical 2 Decarboniza on has only recently become important byproduct of reforma on and represents about 55% of to hydrogen produc on. Increasingly, scholars, environmental ac vists and policy makers have begun to a Coal or petcoke feedstocks yield addi onal CO, roughly twice that classify hydrogen into three broad categories: of natural gas, requiring extra storage. In regions where CO storage resources are limited, this can cause challenges to decarbonizing produc on. BOX 21 Other challenges with hydrogen Hydrogen’s poten al to contribute to decarboniza on has been extensively studied. This focus has revealed several physics and chemistry challenges to widespread hydrogen use. ■ Leakage: Because hydrogen is a very small molecule, leakage risks are substan al, especially in pre-exis ng pipelines or devices. Special materials and gaskets are o en required to ensure minimal leakage. ■ Safety: Hydrogen is colorless, odorless and burns invisibly. On that basis, special monitors and sensors are needed to iden fy opera ng hydrogen combus on units and appropriate mi ga on plans are needed to ensure safety.6 ■ Corrosion and embrittlement: In small frac ons (7-20%), hydrogen can be mixed into exis ng pipeline networks with minimal consequence. At higher frac ons, hydrogen can corrode conven onal pipes, providing a leakage or safety concern. Moreover, hydrogen can make conven onal metal pipes and fi xtures bri le through aging and low-level reac ons. Overt steps are needed to mi gate or counter corrosion and embri lement, poten ally including full pipeline replacement. ■ Storage: Hydrogen is notoriously challenging to store. Many tank systems are adequate (either compressed, liquifi ed or cryo-compressed) but require special materials and systems to avoid leak-off or other losses, some mes adding substan al costs to hydrogen systems. Some work has begun on using engineered salt caverns to store hydrogen in large volumes. While these issues are straigh orward and manageable, they require a en on to ensure safe and cost-eff ec ve hydrogen deployment in industrial se ngs.

December 2019 6 ■ Gray hydrogen: H2 produc on without carbon controls (typically SMR, ven ng

byproduct and combus on related CO2). ■ Blue hydrogen: H2 produc on with carbon controls (typically carbon capture, use and storage—CCUS). ■ Green hydrogen: Electrolysis of water using only low-carbon electricity sources (e.g., renewables, nuclear). All three categories of hydrogen produc on can have a wide range of GHG emissions. For example, both gray and blue hydrogen produc on carry the upstream emissions associated with methane produc on, which can vary substan ally.2,3 Figure 2A-2. Cost of hydrogen produc on ($/kg) of selected hydrogen

For blue hydrogen, CO2 capture can either produc on methods (unsubsidized). Source: Friedmann et al. 2019 be par al (i.e., from only the reforma on unit) or applied to addi onal plant systems at Conven onal produc on from natural gas without addi onal costs (see below).4 Today, four units around carbon capture and storage (CCS)(gray hydrogen) is 5 the world capture CO2 from the reforma on unit, which cheaper than all low-carbon op ons. Par al or full CCS represents an emissions reduc on of roughly 53-60% (blue hydrogen) increases costs by 20-50% depending on per unit hydrogen. It is possible to reduce emissions the degree of decarboniza on. All electroly c hydrogen from hydrogen produc on to much higher levels, is more expensive s ll, with US grid costs producing commonly up to 90%, although even higher capture hydrogen at roughly twice the cost of gray hydrogen and rates are possible. Blue hydrogen produc on is only resul ng in only 20-30% carbon footprint reduc ons.

viable at sites that can access CO2 transport and storage When all power is generated with renewable sources, infrastructure (i.e., where there are geological storage costs increase by a factor of 3-10. sites or pipelines that can move CO2 to them). For green hydrogen, the footprint of electricity A blue-green transition produc on can vary greatly across regions. This Today, it is possible to generate low-carbon hydrogen underscores the need for careful life-cycle analysis at a large scale from natural gas and to decarbonize to understand and es mate the carbon footprint of the produc on with rela vely small increases in cost. all forms of produc on. Many low-carbon electricity It is likely that as CCUS technologies improve, the systems have low capacity factors, which can add incremental cost of decarbonizing blue hydrogen substan al costs for fi rm power genera on. produc on will drop somewhat as well. However, the principal element of blue hydrogen cost is the cost of Estimated costs natural gas itself, which is already low in North America, Following Collidi et al.,7 Friedmann et al.8 developed a and it is hard to imagine drama c cost improvements for “levelized cost of hydrogen” (LCOH). LCOH es mates blue hydrogen. the unit cost of producing hydrogen over its economic In contrast, the primary costs of green hydrogen (low- life me, including capital costs, opera ng and carbon electricity prices) have decreased drama cally maintenance costs, and capacity factors, as well as and con nue to drop. While it is unclear how much calcula ng diff erent costs as a func on of gas costs, costs can or will decrease for solar or wind, it is plausible power costs, conversion methodology and degrees of that capacity factors will increase for some renewable decarboniza on.9 These costs and assump ons are sources and that costs will drop with technology represented in Table 2A-1. and compared in Figure 2A-2. advances.10 While curtailment today represents rela vely small percentages of power genera on, many scholars

December 2019 7 H2 PRODUCTION APPROACH Natural Gas Reforma on* Capture Rate LCOH Cost of Heat (LHV) Steam-methane reforming without CCS 0% $1.05-1.5/kga $8.78-12.51/GJ

53% $1.32-1.77/kg $11.02-14.75/GJ Steam-methane reforming with CCS 64% $1.46-1.91/kg $12.19-15.91/GJ 89% $1.71-2.15/kg $14.22-17.92/GJ Cost of Heat (lower Electrolysis of Water# Cost of Power LCOH hea ng value)

US average grid + PEM (90% capacity factor) $60-90/MWh $4.50-6.04/kg $37.52-50.34/GJ Solar PEV (20% capacity factor) $36-46/MWh $7.1-8.3/kg $59.2-69.2/GJ Wind unsubsidized (35% capacity factor) $29-56/MWh $6.02-7.25/kg $50.17-60.46/GJ

Hydropower unsubsidized $30-60/MWh $4.80-6.34/kg $40.01-52.83/GJ (40% capacity factor)

* All natural gas capture cases assume 90% capacity factor, $3.5/million BTU and $20/ton costs for CO2 compression, transporta on and storage. # All electrolysis cases assume $1,000,000/MW electrolyzer cost. a Even for fi xed gas prices and capacity factors, the range of costs refl ects choice of conversion technology (e.g., SMR vs. autothermal reformers). Table 2A-1: Es mated costs for hydrogen produc on (normalized to natural gas).

have posited that overgenera on of renewables will costs. Improvements could come from novel CO2 prove cost eff ec ve. If so, costs for green hydrogen capture systems that have lower costs themselves, could drop substan ally. They would drop further with from incremental learning-by-doing improvements in substan al capital cost reduc ons for electrolyzers. capital cost (e.g., reduc on in steel, lower cost material subs tu on) or opera ng cost (e.g., improved heat It is thus possible to imagine a transi on from blue to recovery, more effi cient systems). Similarly, for new green hydrogen supply. Low-carbon hydrogen systems hydrogen produc on facili es using conven onal could be deployed rela vely quickly using blue hydrogen technology (e.g., SMRs, ATRs, gasifi ers), costs and as a primary fuel, providing the ability to scale quickly effi ciencies could improve through integrated design. and at modest addi onal cost. Over me, as green hydrogen became cost-compe ve, it could gain market Many groups have studied approaches to improve the share for decarbonized heat and begin to displace blue cost and performance of electrolyzers.9 These include hydrogen produc on. If so, future LCOH could remain discovery and func onaliza on of new materials, most fairly constant and possibly decrease while the total notably metal anodes. Research to improve corrosion frac on of fossil-based hydrogen produc on decreases resistance and seal performance and to extend the over me. capital life and longevity of components and integrated systems remains important. Ul mately, the largest Potential to improve cost element will remain the cost of electric power. To date, only four facili es in the world produce blue Overall research to con nue reducing the total cost hydrogen. It is likely that as more carbon capture for renewable power systems would help. In the near systems are deployed on exis ng facili es, engineers term, research should iden fy and map loca ons and innovators will fi nd opportuni es to decrease where a combina on of features (e.g., high capacity

December 2019 8 BOX 22 Alterna ve approaches to hydrogen produc on Hydrogen can be generated through several other technological pathways. In some cases, these technologies are in early stages of development (low technical readiness level [TRL]). In other cases, the processes are well described and recognized but have not scaled due to high costs or other reasons. ■ Sulfur-iodine cycle: This thermochemical cycle process generates hydrogen from water and recycles sulfur and iodine without their consump on. The cycle operates at high temperatures (~950 °C) from any source, although many consider it to be well-suited to heat from high-temperature nuclear reactors. The Japanese government and Savannah River Na onal Laboratory have studied the process in depth, and a Japanese test reactor runs experiments to improve the effi ciency and performance of the cycle. TRL = 3 ■ Methane cracking: Methane can be separated directly into carbon and hydrogen by breaking its chemical bonds. For example, the Kvaerner Carbon Black & Hydrogen Process (KCB&H) was developed by Norwegian company Kvaerner and uses a high-temperature plasma burner to directly separate methane into hydrogen and amorphous carbon (carbon black). The fi rst plant was built and began opera on in Norway in 1999 but has not received widespread adop on. This process does not emit substan al greenhouse gases, since all carbon is converted to solid form. TRL = 5-6 ■ Biomass gasifi cation: Like natural gas, oil residues, coal or petcoke, biomass can be gasifi ed and combined with water-gas shi to produce hydrogen. This has the advantage of a renewable feedstock (biomass) which could reduce the carbon footprint of produc on drama cally. This process today is expensive due to the high capital costs of gasifi ers,rs, challenges in feed systems and ash handling, and limita ons of biomass supply. There are substan al ranges and uncertain es in carbon footprint. (See Biomass sec on, chapter 2B.) TRL = 8 Because low-carbon hydrogen will remain an important decarboniza on op on for industry and other applica ons (including heavy duty transport), research programs around the world should increase the size and scope of programs to develop new methods of hydrogen produc on.

integrated with system opera on, long-lived programs factors, regular curtailment) produce extremely low-cost would help iden fy possibili es for subs tu on that green power today in proximity to relevant industrial could prove viable. applica ons.

Finally, and perhaps most importantly, research is 1 IRENA, 2018, Hydrogen from renewable power: urgently needed on how best to implement hydrogen Technology outlook for the energy transi on, h ps:// combus on systems in facili es that currently use other www.irena.org/publica ons/2018/Sep/Hydrogen-from- fuels. In some cases, the changes may prove fairly renewable-power modest (e.g., new burner ps, sensors and controls). In 2 IEA, 2019, “Methane emissions from oil and gas”, other cases, subs tu on of low-carbon hydrogen may h ps://www.iea.org/tcep/fuelsupply/methane/ require new handling and fueling systems, as well as new 3 Le Fevre, 2017, “Methane Emissions: from blind spot designs for retrofi ng complex systems and reactors. to spotlight”, The Oxford Ins tute for Energy Studies,

In some cases, addi onal NOX control equipment may h ps://www.oxfordenergy.org/wpcms/wp-content/ be required. For very challenging cases (e.g., cement uploads/2017/07/Methane-Emissions-from-blind-spot- kilns or blast furnaces) where solid fuel use is closely to-spotlight-NG-122.pdf

December 2019 9 4 Soltani R, Rosen M.A., Dincer I., 2014, Assessment 8 J. Friedmann et al., 2019, “Low-carbon heat solu ons of CO capture op ons from various points in steam for heavy industry: sources, op ons & costs today”, methane reforming for hydrogen produc on, Columbia Univ. Center on Global Energy Policy Intl. J. of Hydrogen Energy 39(35) DOI: 10.1016/j. 9 DOE, 2015, Quadrennial Technology Review, Chapter ijhydene.2014.09.161 6, Innova ng Clean Energy Technologies in Advanced 5 GCCSI, 2019, Global Status Report, h ps://www. Manufacturing, h ps://www.energy.gov/sites/prod/ globalccsins tute.com/resources/global-status-report/ fi les/2016/06/f32/QTR2015-6I-Process-Hea ng.pdf 6 NREL, 2008, Hydrogen Conversion Factors and Fact 10 Gigler J, Weeda M, 2018, Outlines of a Hydrogen Card, DOE/GO-102008-2597, h ps://www.nrel.gov/ Roadmap, TKI Nieuw Gas, h ps://www. docs/gen/fy08/43061.pdf topsectorenergie.nl/sites/default/fi les/uploads/ TKI%20Gas/publica es/20180514%20Roadmap%20 7 G. Collidi, et al., 2017, “Techno-Economic Evalua on Hydrogen%20TKI%20Nieuw%20Gas%20May%202018. of Deploying CCS in SMR Based Merchant H pdf Produc on with NG as Feedstock and Fuel”, Energy Procedia, 114:2690-2712, h ps://doi.org/10.1016/j. egypro.2017.03.1533

December 2019 10 BIOMASS grassland to plant energy crops can result in payback 3 Biomass is the oldest source of industrial heat and periods of decades or centuries. provides about 10% of global primary energy.1 Of this, Complica ng the picture, many energy crops, including roughly 15% goes to non-electricity industrial uses (7.8 corn and sugarcane, can compete with food crops for EJ in 2009) with the rest used mainly for cooking, space land, which has ripple eff ects on the food system and can hea ng, vehicle fuel and electricity genera on.2 result in indirect land-use changes with carbon impacts. Expanded use of biomass has substan al poten al to The lifecycle impacts, including the indirect land-use decarbonize industrial heat produc on. A wide variety of changes, have been studied extensively, primarily in the biomass types, from wastes such as manure and almond context of liquid biofuels for transporta on. Some recent shells to dedicated energy crops such as sugarcane research has indicated that the CO2-emissions impacts and switchgrass, can be used to provide industrial of indirect land-use changes may not be as strong as 4 heat. Exis ng or emerging technologies are available previously thought. However, lifecycle impacts are s ll to convert biomass to many intermediates, such as signifi cant, especially for liquid fuels derived from energy biomethane, biodiesel or bio-char. Unlike electricity crops. Figure 2B-1 summarizes es mates of the lifecycle or hydrogen, biomass can act as a carbon source and impacts from biofuels and fossil fuels. The greenhouse chemical reductant in steelmaking and other processes. gas emissions impact of genera ng heat from biomass is generally low, with es mates ranging from negligible 3,5 Climate impacts of bioenergy to 30 g CO2eq/MJ. For comparison, the impacts range roughly 80-150 g CO eq/MJ for fossil fuels, as shown in Although combus on of biomass releases similar 2 Figure 2B-1. quan es of CO2 as fossil fuels, biomass combus on can have substan ally lower climate impacts because the Current industrial use of bioenergy is dominated by solids (93% in the EU), followed by municipal wastes CO2 will be recaptured from the atmosphere when the source of the biomass regrows. For example, combus on (3%) and biogas (2%).6 To the extent that industrial use of agricultural residues, like rice husks, is generally of bioenergy con nues to favor solids and gases over considered carbon-neutral because the crops will regrow liquids, which is reasonable given current demands the next season. This is also true for short-rota on and expected sources of biomass, the lifecycle impacts woody biomass from sustainably managed forests. are somewhat less challenging for industry than for However, use of biomass can result in net land-use transporta on. S ll, most use cases of biomass for industrial heat are not truly carbon neutral. However, changes, reducing or elimina ng any CO2 emissions benefi ts. Transport, processing and use of fer lizer can bioenergy could be made carbon neutral or carbon nega ve by addi on of carbon capture and storage (CCS) also result in CO2 emissions, reducing or elimina ng the to either the industrial process or the biofuel processing CO2 benefi ts of bioenergy. facility, as discussed below. If supply of dedicated energy crops is substan ally expanded, some land must be converted from other Biomass availability uses to grow the crops. There is typically a net change in carbon stored in the soil and fl ora when land use Numerous assessments have been made of global changes. If the previous use stored more carbon than biomass availability. Es mates vary widely, especially the energy crop, then there will be an ini al release of for dedicated energy crops, but there is moderately carbon to the atmosphere, followed by a reduc on in good agreement in the literature that 200-500 EJ/y carbon emissions to the atmosphere, compared to the of sustainably produced biomass can be available 3,5 scenario where fossil energy was used instead. Thus by 2050. These values compare favorably with the a “payback period” can be calculated based on the es mated global industrial energy demand of 230 EJ 7 number of years it would take to compensate for ini al in 2012, projected to rise to 330 EJ by 2040. However, carbon release. When conver ng degraded land or other there are compe ng demands for biomass in a low- types of cropland to energy crops, the payback period carbon future, including as vehicle fuel, dispatchable is generally a year or less. But clearing forest or na ve electricity and a means of nega ve emissions through bioenergy with CCS (BECCS). Most assessments of

December 2019 11 Figure 2B-1. Summary of literature es mates of the lifecycle greenhouse gas impacts of bioenergy. Source: IPCC3 (used with permission).

bioenergy as a climate-mi ga on op on have focused wider range of uses. Common pathways are summarized on liquid biofuels for transporta on specifi cally. These in Figure 2B-2. compe ng uses will have to be balanced in a policy that Most biomass sources, such as forest biomass and encourages expanded use of biofuels in industry. agricultural residue, contain up to about 50% water. Also limi ng its deployment in industry, biomass is more Chipping and drying is the most common treatment geographically diverse and more expensive to collect and for biomass currently used industrially, and this transport than fossil fuels. For example, in Australia, bio- allows reasonable transport. There is already robust char suitable for steelmaking was found to cost about 4 interna onal trade in wood, which is the 5th most mes as much as coal, even though low-cost agricultural important traded commodity. Only about 10% of residues are also abundant there.8 The produc on cost currently traded woodchip goes for bioenergy, but s ll for solid biomass has been es mated to be in the range pelle zed wood for energy is traded interna onally, of 11-50 $/GJ using current technologies,3 which makes primarily in Europe. With emerging technologies, these it at least four mes as expensive as coal and twice as sources can be treated by a handful of other means expensive as natural gas in current markets.9,10 This to produce fuels for transport or further processing: suggests that much of the poten al biomass discussed gasifi ca on to produce biogas, syngas or hydrogen; above will be uneconomical to collect without strong pyrolysis to produce bio-oil and bio-char; torrefac on incen ves for industrial decarboniza on. to produce torrefi ed biomass; and hydrothermal liquefac on to produce bio-oil. Processing and transport As examples, top-level process diagrams for biomass Since biomass grows over wide stretches of land, gasifi ca on to hydrogen (Figure 2B-3) and biomass collec on and transport is o en a crucial component pyrolysis (Figure 2B-4) are shown below. With of the cost to use it. Woody biomass has about half gasifi ca on, a range of biomass types can be converted the energy density and considerably lower bulk density into renewable hydrogen, which in turn can provide (before grinding) than coal.11 However, biomass can be carbon-free heat. Depending on the biomass source,

converted to a variety of forms for easier transport and a the CO2 emissions resul ng from the process may be

December 2019 12 Figure 2B-2. Biomass conversion pathways. (Fire Management refers to forest biomass, such as small-diameter trees and shrubs, removed from a forest to reduce fi re risk or fi re severity. RNG (renewable natural gas) refers to biogas that has been purifi ed to meet natural gas pipeline standards (also known as biomethane).

considered carbon neutral. However, this process is diges on is the least expensive source of bioenergy, with especially a rac ve because the rela vely pure stream produc on costs es mated in the range of 1.5-8.7 $/GJ, 3 of CO2 from the acid-gas-removal opera on can be some mes less than fossil gas. However, the feedstocks captured and sequestered, resul ng in nega ve CO2 for biogas are limited: manure, MSW, landfi ll gas and emissions (analogous to BECCS), while also providing a agricultural waste made up about 10% of total bioenergy stream of valuable hydrogen. use in 2007,1 and these are unlikely to grow as much with demand as other types of biomass. In the pyrolysis process, biomass is heated without oxygen to frac onate the material into gases, liquid fuels For most of the conversion pathways, addi onal refi ning and bio-char. The bio-char and liquids can be used as can yield a fuel that is a drop-in replacement for a fossil industrial fuels. Typically, the gases are combusted to fuel. Bio-char, depending on the process and biomass

produce heat for the process, and the resul ng CO2 can source, has a hea ng value of 30-38 MJ/kg, which is in 11 op onally be captured to yield reduced or nega ve CO2 the range of many coals. Anaerobic diges on generates

emissions for the overall process. a biogas that is roughly half methane and half CO2. If the CO and minor impuri es are removed, the remaining Wet biomass streams, such as manure, wastewater, 2 biomethane can meet specifi ca ons for natural gas landfi lled municipal solid waste (MSW) and food pipelines. Syngas can be converted to desired fuels waste are most commonly treated with anaerobic using conven onal Fischer-Tropsch methods. Pyrolysis diges on, crea ng biogas. In some cases, hydrothermal oil, although it starts with about half the hea ng value liquefac on can also be used. Biogas via anaerobic of crude oil because of its high oxygen content, can

Figure 2B-3. Process for biomass gasifi ca on to hydrogen.

December 2019 13 Figure 2B-4. Process for biomass pyrolysis to liquid fuel and bio-char.

be refi ned to standard vehicle fuels, depending on the be er func on in exis ng steel mills and development desired product. of the basic oxygen furnace process to run be er on biomass fuels are both important pathways to In most cases, refi ned biomass fuels are more expensive decarbonize steelmaking. than compe ng fossil fuels, which has limited the market for conversion technologies in the same way as for Regional biomass availability appears less of a concern biomass use. Except for anaerobic diges on, which is for steelmaking. In an analysis of biomass availability rela vely mature, technology development and market compared with the loca ons of the current steelmaking incen ves can s ll substan ally improve the conversion industry, it was found that several of the top steel- processes and lower costs. producing countries have high suitability to adopt biofuels in steelmaking, including China, Russia, the US Specific applications and Brazil. Japan, India and Germany had moderate For certain industrial uses where the process is sensi ve suitability. Only South Korea, Ukraine and Taiwan had to fuel composi on, further development is especially low suitability to biomass adop on, together accoun ng 8 needed. As discussed in Chapter 3B, steelmaking is a for 6.6% of global steel produc on. complex process with both fuel and process emissions In contrast to steel mills, cement kilns are fairly tolerant coming from mul ple units of an integrated plant. to fuel varia ons and impuri es because of their high Biomass can be subs tuted in several forms. Bio-coke combus on temperature. Although not typical prac ce, can be produced to replace coal-based coke in the many cement plants across the world co-fi re wastes and coking opera on. Bio-char can be used in the sintering biomass along with fossil fuels when local condi ons process and blast furnace. Bio-chars with higher hea ng make this economically a rac ve.13 Addi onally, co-fi ring values are more eff ec ve in the blast furnace. Pyrolyzed can have benefi ts for local air pollu on by reducing or torrifi ed biomass may also be op ons to fuel the blast emissions of sulfur and nitrogen oxides. Expanded use of furnace.12 biomass in cement opera ons appears straigh orward, Mul ple full-scale steelmaking plants already operate however it will not address the process CO2 emissions on biomass energy in Brazil, with 34% of the energy from calcium carbonate decomposi on. consumed by the iron and steel industry in that country coming from biomass.6 Elsewhere, however, only low 1 WEC, 2010 Survey of Energy Resources, World Energy subs tu on rates of biomass have generally been Council, London, UK, 2010. achieved.8 The composi on and proper es of bio-coke, 2 Esa Vakkilainen, Katja Kuparinen, Jussi Heinimö, Large in par cular, need refi nement and innova on—bio-coke Industrial Users of Energy Biomass, IEA Bioenergy Task does not yet perform as well as conven onal coke due 40, 2013. to higher reac vity and lower strength a er reac on. 3 Helena Chum, Andre Faaij, José Moreira, Göran In general, development of the biomass feedstocks to Berndes, Parveen Dhamija, Hongmin Dong, Benoît

December 2019 14 Gabrielle, Alison Goss Eng, Wolfgang Lucht, Maxwell 8 H. Mandova, W.F. Gale, A. Williams, A.L. Heyes, P. Mapako, Omar Masera Ceru , Terry McIntyre, Tomoaki Hodgson, K.H. Miah, Global assessment of biomass Minowa, Kim Pingoud, Bioenergy, in: IPCC Spec. Rep. suitability for ironmaking – Opportuni es for Renew. Energy Sources Clim. Change Mi g., Cambridge co-loca on of sustainable biomass, iron and steel University Press, Cambridge, UK, 2011. h ps://www. produc on and suppor ve policies, Sustain. Energy ipcc.ch/report/renewable-energy-sources-and-climate- Technol. Assess. 27 (2018) 23–39. doi:10.1016/j. change-mi ga on/. seta.2018.03.001. 4 Glaucia Mendes Souza, Reynaldo L. Victoria, Luciano 9 EIA, Natural gas prices, (n.d.). (accessed September M. Verdade, Carlos A. Joly, Paulo Eduardo Artaxo 21, 2019). Ne o, Carlos Henrique de Brito Cruz, Heitor Cantarella, 10 EIA, Coal Markets, (n.d.). h ps://www.eia.gov/coal/ Helena L. Chum, Luis Augusto Barbosa Cortez, Rocio markets/ (accessed September 21, 2019). Diaz-Chavez, Erick Fernandes, et al., SCOPE: Bioenergy and Sustainability, UNESCO, 2014. 11 K. Raveendran, A. Ganesh, Hea ng value of biomass and biomass pyrolysis products, Fuel. 75 (1996) 5 F. Creutzig, N.H. Ravindranath, G. Berndes, S. Bolwig, R. 1715–1720. doi:10.1016/S0016-2361(96)00158-5. Bright, F. Cherubini, H. Chum, E. Corbera, M. Delucchi, A. Faaij, J. Fargione, H. Haberl, G. Heath, O. Lucon, R. 12 E. Mousa, C. Wang, J. Riesbeck, M. Larsson, Biomass Plevin, A. Popp, C. Robledo-Abad, S. Rose, P. Smith, A. applica ons in iron and steel industry: An overview Stromman, S. Suh, O. Masera, Bioenergy and climate of challenges and opportuni es, Renew. Sustain. change mi ga on: an assessment, GCB Bioenergy. 7 Energy Rev. 65 (2016) 1247–1266. doi:10.1016/j. (2015) 916–944. doi:10.1111/gcbb.12205. rser.2016.07.061. 6 I. Malico, R. Nepomuceno Pereira, A.C. Gonçalves, 13 A. Rahman, M.G. Rasul, M.M.K. Khan, S. Sharma, A.M.O. Sousa, Current status and future perspec ves Recent development on the uses of alterna ve fuels for energy produc on from solid biomass in the in cement manufacturing process, Fuel. 145 (2015) European industry, Renew. Sustain. Energy Rev. 112 84–99. doi:10.1016/j.fuel.2014.12.029. (2019) 960–977. doi:10.1016/j.rser.2019.06.022. 7 EIA, Chapter 7: Industrial sector energy consump on, in: Int. Energy Outlook 2016, U.S. Energy Informa on Administra on, 2017.

December 2019 15 ELECTRIFICATION distribu on grid build-out and transformer installa on) and place severe demands on the local electric grid Overview (see below). In markets where natural gas is cheap and A wide variety of electrical technologies are available electricity is expensive, electrical hea ng systems will to provide industrial process heat. When powered by be at a cost disadvantage. Also, industrial processes that low-carbon electricity, these technologies can provide are op mized for combus on-based process heat are process heat with very low greenhouse gas (GHG) o en highly op mized, taking advantage of waste heat emissions. Although the opera onal principles and for combined heat and power or recupera ng it through capabili es of these technologies vary, they share heat exchangers or heat pumps. In these cases, replacing several features because of their use of electricity. combus on heat with electrical hea ng may require substan al plant redesign. Unlike approaches to low- First, electric hea ng systems can be precisely controlled carbon heat treatment that are essen ally low-carbon by varying the electrical current, voltage or other fuel replacements, such as hydrogen and biofuels, parameters. This allows precision delivery of heat, electrical hea ng cannot leverage exis ng combus on- minimizing energy waste and enabling automated based industrial process designs and generally cannot process control. This feature stands in contrast to be retrofi ted into an exis ng process without signifi cant combus on-based process heat, for which the amount equipment modifi ca on. of heat delivered is more diffi cult to control. Similarly, electrical systems are able to provide heat fl exibly at Principles of electrical heating a range of design temperatures. Combus on systems, in contrast, are constrained by the combus on At the most basic level, heat can be transferred to a temperatures of their fuels. process material in three ways. Convec ve hea ng is the transfer of heat energy through the mo on of a A second, related feature of electrical hea ng systems is fl uid, such as water or air; it is generally constrained by their ability to rapidly turn on and off . This allows them parameters such as fl uid fl ow rate and heat capacity. to have greater opera onal fl exibility, ramping up and Conduc ve hea ng results from direct contact between down heat delivery for a range of purposes. These can the process material and a solid heat source; it is include adjus ng processing opera ons based on grid or impacted by the thermal conduc vity of the process market condi ons or enabling extremely rapid, brief heat material. Radia ve hea ng is caused by electromagne c applica on unit opera ons compared with combus on (EM) waves (such as microwaves) arriving at the process methods (for example, electron-beam curing). material; it is constrained by both the refl ec vity and Third, electrical hea ng systems tend to have rela vely absorp on of the process material to the wavelengths low maintenance. They are not exposed to combus on used. products or fl ame, and their components are almost In an engineering context, there are two broad en rely solid-state, with no fuel supply or storage approaches to applying heat energy. Direct hea ng requirement. These features tend to reduce or applies an electric current through the process material eliminate problems such as corrosion from combus on to cause resis ve hea ng, induces an electric current gases or fl ame impingement on refractory materials. in the process material using alterna ng magne c Some electrical hea ng systems, such as induc ve fi elds, or excites molecules within the process material and microwave technologies, can apply heat without with electromagne c radia on (microwaves or radio contac ng the workpiece or material being treated, frequency). In each of these cases the material must which reduces the poten al for contamina on and have suitable proper es (such as electrical conduc vity). enables be er control of the reac on environment Indirect hea ng is used in cases where the process during heat applica on. material is not suitable for direct hea ng and instead The disadvantages of electrical hea ng systems include uses one of these methods to heat a separate susceptor the need to provide large amounts of electric power, or element that is near or in contact with the process which may require addi onal infrastructure (such as material, which then transfers that heat to the process material through conduc on, convec on or radia on.

December 2019 16 Figure 2C-1. An infrared dryer for automobile paint. Short-wave infrared emi ers can reach fi lament temperatures of 2,000°C. (Dmitry Kalinovsky/Shu erstock.com)

Specific electrical heating methods Infrared hea ng is based on passing an electric current through a solid resistor to heat it and then direc ng the Direct resistance hea ng is the simplest of all electrical resul ng infrared radia on to the process material. The hea ng methods, par cularly for conduc ve material material must have rela vely high absorp on and low that can be directly heated by the applica on of electric refl ectance for infrared wavelengths corresponding to current. The material can be clamped to electrodes in the temperature of the radiator. Short-wave emi ers the wall of the furnace in order to apply current. Joule reach the highest temperatures, up to approximately hea ng resul ng from the interac on of the current 2,000 °C, using evacuated quartz tubes with tungsten and the electrical resistance of the process material can fi laments, back-fi lled with argon to prevent oxida on. be highly effi cient, par cularly for materials with high Medium- and long-wave emi ers use the tubular resistance (e.g., steel). Temperatures up to 2,000 °C are hea ng elements described above or wires embedded in possible.1,2 ceramic panels. Baffl es and refl ectors can focus infrared Indirect resistance hea ng uses electrical resistance radia on on the process material, improving effi ciency. in a hea ng element, which is commonly made from The heat transfer is mostly confi ned to the surface of the graphite, silicon carbide or nichrome (nickel-chromium process material and is therefore most appropriate for alloy). The heat is transferred by conduc on, convec on surface applica ons like curing and drying.1,2 or radia on (similar to infrared hea ng, below) to the Microwave hea ng is based on the fact that microwave work material. Various geometries are used; tubular radia on (with frequencies in the range of 300 to hea ng elements are common and comprise a nichrome 300,000 MHz) heats non-conduc ve (dielectric) hea ng coil surrounded by magnesium oxide for materials that are composed of or contain polar electrical insula on, sheathed in stainless steel. These molecules, such as water. Microwaves excite these systems have a maximum temperature of approximately molecules into mo on, which leads to fric on hea ng. 750 °C and can deliver heat at powers ranging up to 120 The heat energy can be deposited throughout the bulk Wa s per square inch of surface area. Indirect resistance of the process material as long as it is not too thick. hea ng is used in electric indirect rotary kiln technology, However, because microwave radia on is coherent with in which resis ve heaters are placed outside of a rota ng a wavelength in the range of a millimeter to a meter, high-temperature alloy shell and heat is transferred standing waves can develop in hea ng chambers, leading to process material inside; these systems can reach to local hot and cold spots. Microwaves are generated in temperatures of 1,200 °C.1,2

December 2019 17 a magnetron and generally must be guided or contained material. This deposits heat both from the direct to ensure effi ciency and minimize exposure. Radio resistance of current passing through the material and frequency hea ng works on a very similar principle, from the radiant energy from the arc. Commercialized although it uses lower frequencies (2 to 100 MHz) with arc furnaces range from a few tons to hundreds of correspondingly longer wavelengths. These are able to tons of capacity. The electrodes wear out and must be penetrate farther into process materials, although they replaced; suitable electrode materials are important tend to deliver heat more slowly.2,3 for the overall economic viability of the technology. An alternate but less common confi gura on is the indirect Induc on hea ng is based on an alterna ng magne c arc furnace, which draws the arc between electrodes, fi eld, generated by passing an AC electric current applying heat through radiant transfer. Electric arc through a coil (solenoid). This fi eld in turn induces furnaces are commonly used in steelmaking, where alterna ng eddy currents in the work material if it is they achieve temperatures up to 1,800 °C, as well as in electrically conduc ng. For op mal effi ciency, the work the produc on of ferronickel in the Rotary Kiln–Electric material is placed within the solenoid, or the magne c Furnace (RKEF) process.2,4,5 fl ux is coupled into the material in other ways. Induc on hea ng avoids any physical contact between the hea ng Electron beam hea ng uses a focused beam of system and the process material. However, the energy is electrons directed onto a process material, usually in mostly deposited on the surface of the material (due to vacuum. Common uses include cross-linking polymers, the skin eff ect). If the work material is not conduc ng, it welding, surface hardening for high-wear automo ve can be put in contact with a susceptor, which is heated components, and addi ve manufacturing. Electron beam induc vely and transfers heat through conduc on or furnaces are used in mel ng refractory metals such as convec on. Common applica ons are refi ning and tanium. The hea ng is primarily at the surface of the re-mel ng of metals, including aluminum, copper, brass, material, making bulk treatment more challenging.6,7 bronze, iron, steel and zinc.1,2 Plasma heaters operate by developing an electric arc Electric arc furnaces consist of a refractory vessel with across two cooled electrodes; gas (of many diff erent retractable electrodes, o en made from graphite composi ons, including a variety of waste gases) is or carbon. AC or DC current is passed through the directed past the arc, which ionizes it into plasma that electrodes and forms an electric arc with the process can reach temperatures from 2,000–20,000 °C. The

Figure 2C-2. Electric arc furnaces use electric current to form an arc between electrodes, providing high-temperature heat to melt scrap steel and iron. (D.Alimkin/Shu erstock.com)

December 2019 18 plasma forms a jet, which is then directed onto the work redesigns. As a consequence, the capital costs for material, hea ng it. Plasma processing is commonly used process changes to electrifi ca on are generally higher in the tanium industry, as well as in the disposal of toxic than those for switching to alterna ve fuels. ash, asbestos and sludge.8,9 For low-temperature process hea ng requirements System consideration (generally under 200 °C) there are several electrical Electrifi ca on of process heat can create signifi cant technologies available that make use of waste heat, diffi cul es for local electric-grid opera on. Large power including heat pumps and organic Rankine cycle consumers that func on in a batch mode are par cularly turbines. Solar process hea ng is also available for challenging, since they can rapidly increase or decrease temperatures up to 250 °C.10 However, these are not overall power demand and require genera on to ramp eff ec ve for medium- or high-temperature process heat. quickly. In some cases, such as electric arc furnaces for steel produc on, there may be strategies to harmonize Installation considerations opera ons with demand-side management (DSM) Combus on-based hea ng for a range of industrial systems put in place by grid operators, but this can 11,12 processes usually involves furnaces whose design lead to highly complicated ming decisions. In other has been op mized for delivering heat from one or cases, such as electroly c produc on of hydrogen, it more individual point loca ons (burners) where fuel is may be possible to add fl exibility to a con nuous process 13 combusted. The furnace is designed to handle the fl ow to par cipate in DSM or provide other grid services. of combus on and reac on gases and may also include Ul mately, wide-scale electrifi ca on of process heat heat integra on to recapture waste heat for other uses would require more integrated system planning between in the overall process. While some modifi ca ons are industrial customers and grid operators to be er required for this design to burn alterna ve fuels such understand the opportuni es and challenges. as hydrogen and biomass, the basic architecture largely remains the same. However, modifying the design to use Conclusions electric sources of heat requires much more substan al ■ A wide variety of electrical technologies are available changes. for delivering process heat. These can achieve temperatures of well over 1,000 °C and, in some cases, ■ First, electric heat is generally not delivered from point over 10,000 °C. sources (burners). Direct hea ng methods deposit ■ heat energy throughout a material or poten ally in a Electrifi ed process hea ng has several advantages surface layer. Indirect hea ng generally delivers heat over combus on-based hea ng, including precision across the surface of a material. This can signifi cantly temperature control, high fl exibility and low change the distribu on of temperatures within maintenance costs. However, it requires large amounts a furnace and thus the hea ng rates of the work of electric power, which may be unavailable or cost material, poten ally requiring process redesigns. prohibi ve. ■ ■ Second, there is no need to handle combus on Electric hea ng systems can be categorized as direct, gases, as none are generated in the hea ng process. in which a working material is heated directly through However, heat-integra on systems will no longer be electric resistance, microwaves or other techniques, able to use this waste heat, poten ally leading to a or as indirect, in which a separate device such as a cascade of necessary process changes throughout the susceptor or resistor is used to deliver heat through overall facility. conduc on, convec on or radia on. ■ ■ Third, replacing fuel combus on with electrifi ca on Installa on of electric hea ng systems in facili es removes the need to handle fuel delivery to burners, currently using combus on-based hea ng may require but it may create the need for managing high-voltage substan al process changes and corresponding capital electric power distribu on through ac ve cooling, investment. It may also eliminate heat integra on electrical isola on, etc. In general, the architecture methods that had been in use to recuperate waste and design assump ons of electric process heat are heat, crea ng requirements for installa on of very diff erent from those for fuel-combus on-based addi onal hea ng systems. process heat and lead to a far larger need for process ■ Electrifi ca on of process heat may also place

December 2019 19 signifi cant strain on the local electric grid, requiring 6 M. Gala , et al. 2018, “A literature review of capital-intensive upgrades to electric transmission and powder-based electron beam mel ng focusing on distribu on facili es. In some cases these systems can numerical simula ons”, Add. Man. 19:1-20, h ps://doi. org/10.1016/j.addma.2017.11.001. par cipate in demand-reduc on programs to ease the impact on the grid, but in others this may not be 7 N. Venkatramani, 2002, “Industrial plasma torches and possible due to the need for con nuous or infl exible applica ons”, Cur. Sci., 83(3):254-262, h ps://www. process heat delivery. jstor.org/stable/24106883?seq=1#page_scan_tab_ contents 8 L. Rao, et al., 2013, “Thermal Plasma Torches for 1 US DOE, “Improving process hea ng system Metallurgical Applica ons”, In: 4th Int. Symp. H. Temp. performance: A sourcebook for industry” (2007), Met. Proc., TMS Mee ng, San Antonio, TX, h ps://doi. h ps://www.energy.gov/sites/prod/fi les/2014/05/f15/ org/10.1002/9781118663448.ch8. process_hea ng_sourcebook2.pdf. 9 A. Sharma, et al., “2017 Solar industrial process 2 McMillan, C. et al. “Genera on and use of thermal hea ng: A review.” Renewable and Sustainable Energy energy in the U.S. industrial sector and opportuni es to Reviews, 78:124-137, h ps://doi.org/10.1016/j. reduce its carbon emissions”, NREL/INL (2016), h ps:// rser.2017.04.079. www.nrel.gov/docs/fy17os /66763.pdf. 10 Zhang, X. et al. “Cost-eff ec ve scheduling of steel 3 Leonelli, C. et al, 2010, “Microwave and ultrasonic plants with fl exible EAFs”, IEEE Transac ons on processing: Now a realis c op on for industry”, Chem. Smart Grid (2016), h ps://ieeexplore.ieee.org/ Eng. Proc.: Proc. Int. 49(9):885-900, h ps://doi. document/7482812. org/10.1016/j.cep.2010.05.006. 11 Ave, G. et al. “Demand side management scheduling 4 G. Li, et al. 2016, “Ferronickel prepara on from formula on for a steel plant considering electrode nickeliferous laterite by rotary kiln-electric arc furnace degrada on”, IFAC-PapersOnline (2019), h ps:// process” In: S. Ikhmayies, et al. (eds.) Characteriza on www.sciencedirect.com/science/ar cle/pii/ of Minerals, Metals, and Materials 2016. Springer, S2405896319302307. Cham, h ps://link.springer.com/content/ pdf/10.1007%2F978-3-319-48210-1_17.pdf. 12 Brolin, M. et al. “Industry’s electrifi ca on and role in the future electricity system”, Chalmers B. Lee, B. et al., 2014, “Review of Innova ve Energy University (2017), h ps://research.chalmers.se/en/ Savings Technology for the Electric Arc Furnace”, JOM, publica on/248528. 66(9):1581-1594. h ps://doi.org/10.1007/s11837-014- 1092-y 5 Industrial Hea ng, 2001, “Advancements In Electron Beam Mel ng and Refi ning”, h ps://www. industrialhea ng.com/ar cles/84954-advancements-in- electron-beam-mel ng-and-refi ning.

December 2019 20 CARBON CAPTURE, USE AND STORAGE pressure of the CO2 and H2 mixture is rela vely high, as is Carbon capture, use and storage (CCUS) is a collec on of the concentra on of CO2 therein, making the separa on process less energe cally intensive and more compact. technologies that result in substan al reduc ons of CO2 emissions.1,2 The building blocks of CCUS include: Solvent-based separa ons dominate, but membrane and adsorp on processes (e.g., pressure swing adsorp on) ■ separa on of CO2 from combus on products (such as are also increasingly popular in commercial applica ons. fl ue gas) or hydrocarbon fuels; In principle, this is the route used for CO2 capture from ■ transporta on of CO2 to a suitable geologic storage steam-methane reforming (SMR) in produc on of blue site; hydrogen—although, it may also be combined with ■ injec on of CO2 into a reservoir where it becomes post-combus on capture to maximize CO2 removals trapped deep underground; and/or from SMR.4 Pre-combus on capture for CCUS is used ■ 5 6 use of CO2 in enhanced oil recovery, alkaline minerals commercially today in fer lizer produc on, refi ning and (e.g., steel slag), aggregates, chemicals, fuels or other SMR.7,8 products.3 Post-combus on capture is separa on of CO from the CCUS does not, strictly speaking, decarbonize produc on products of combus on—referred to commonly as fl ue of industrial heat—a er all, CCUS is based on the gas (Figure 2D-2). Flue gas is composed primarily of nitrogen, CO, water and lesser amounts of pollutants capture of CO2 that results from use of carbon- containing fuels. However, CCUS is an a rac ve op on (e.g., oxides of sulfur and nitrogen), where the nitrogen to reduce emissions from industrial heat because CCUS comes from the atmosphere. Flue gas is typically at does not, in principle, require a wholesale change in the close to atmospheric pressure, with a rela vely low CO design of industrial facili es or underlying produc on concentra on (i.e., typically less than 10% by volume), processes. However, as the case studies on cement, iron making the separa on typically more energy intensive and steel, and chemicals in this Roadmap illustrate, CCUS than for pre-combus on capture. The clear benefi t provides maximum benefi t when closely integrated of post-combus on capture, however, is that it can with underlying industrial processes. There are generally be added as an “end-of-pipe” solu on for almost any sta onary combus on process using any fuel. Solvent- three routes to CO2 capture, all of which have relevance in industrial applica ons: pre-combus on, post- based absorp on processes dominate in commercial combus on and oxy-combus on. applica ons, but advanced solvents, adsorbents and other processes (e.g., calcium looping) are being In pre-combus on CO capture, a hydrocarbon fuel is 2 developed.2,9 Post-combus on capture for CCS has been converted to a mixture composed predominantly of commercially applied in power genera on in two cases.5 CO2, hydrogen (H2) and water from which the CO2 is separated, leaving hydrogen for use as a fuel (Figure The third route to CO2 capture is referred to as oxy- 2D-1). The hydrocarbon fuel can be natural gas or a solid combus on (Figure 2D-3). In its most straigh orward fuel—e.g., coal, biomass—that has been gasifi ed. The implementa on, a hydrocarbon fuel is burnt in oxygen— advantage of pre-combus on capture is that the total typically diluted with recycled CO2 for temperature

Figure 2D-1. Typical pre-combus on capture process for solid fuels (coal, biomass) and natural gas. The water-gas shi reac on is not illustrated here for the sake of simplicity.

December 2019 21 Figure 2D-2. Typical post-combus on capture process for coal, biomass and natural gas. Cleaning of the fl ue gas is not illustrated here for the sake of simplicity.

control. The resul ng combus on products are CO2 large-scale CCS. Na onal and interna onal standards and water, the la er of which can be easily removed, exist for design and construc on of such pipelinesa, as 10 leaving CO2. Chemical looping processes can also be do government safety regula ons where such pipelines b considered an oxy-fuel route, but chemical looping exist . Shipping of CO2 is also currently prac ced at small combus on may be less relevant to industrial processes scales and is being considered as part of a Norwegian than the closely related chemical looping reforming, integrated CCS demonstra on projectc. Transport by rail which could be an alterna ve means to produce or truck may also be an economically viable op on for 11 hydrogen while capturing CO2. Similarly, power- small-scale sources (e.g., less that 100 ktCO2/y) over genera on cycles based on oxy-combus on of natural rela vely short distances.13 gas are in development,10 but these typically are less Experience with CO storage in geological forma ons relevant for industrial-process heat applica ons. The 2 has been growing, as EOR projects that inject and store notable diff erence between the oxy-combus on routes CO have been undertaken since the 1960s and the fi rst and others is that oxy-combus on does not require CO 2 2 dedicated geological storage project began opera ons separa on and instead involves separa on of oxygen in 1996. In addi on, governments have con nued to from air. Oxy-fuel CO capture has been demonstrated 2 support research to advance tools and methods for at industrially relevant scales but has not yet been measuring and predic ng the behavior of stored CO . commercially applied. 2 This growing knowledge base has been refl ected in

Once CO2 has been captured from an industrial process development of standards for geological storage in 14,15 via one (or more) of these three routes, the CO2 must be recent years. transported to a loca on suitable for geologic storage A suitable geologic storage site must have suffi cient and then injected into the storage reservoir. Perhaps capacity to hold the desired quan ty of CO , while the most well-prac ced component of the CCUS chain 2 also being able receive the CO2 at acceptable rates is CO2 transport. In the US, over 7,000 km of pipeline transport around 70 MtCO/y of CO2, predominantly for a use in enhanced oil recovery (EOR).12 The US pipeline E.g., in Canada, CSA Z662; Europe, DNV-RP-J202; and, interna onally, ISO 27913:2016. network extends into Canada as well, delivering CO2 for b E.g., in the United States, 49 CFR Part 195. use in EOR, and is being extended in Alberta to enable c See h ps://ccsnorway.com/

Figure 2D-3. Typical oxy-combus on process for solid fuels (e.g., coal, biomass). Natural gas could also be used in such a process.

December 2019 22 through a reasonable number of wells. It must also Capture and Storage, Cambridge University Press, Cambridge, U.K., 2005. safely contain this CO2 permanently—or at least for the 16 foreseeable future. mainly coal, for power genera on 2 M. Bui, C. S. Adjiman, A. Bardow, E. J. Anthony, A. and combus on in industrial processes because they Boston, S. Brown, P. S. Fennell, S. Fuss, A. Galindo, are rela vely abundant, cheap, available and globally L. A. Hacke , J. P. Halle , H. J. Herzog, G. Jackson, J. distributed, thus enhancing the security and stability Kemper, S. Krevor, G. C. Maitland, M. Matuszewski, I. of energy systems. Geological media suitable for CO S. Metcalfe, C. Pe t, G. Puxty, J. Reimer, D. M. Reiner, E. S. Rubin, S. A. Sco , N. Shah, B. Smit, J. P. M. Trusler, storage through various physical and chemical trapping P. Webley, J. Wilcox and N. M. Dowell, Energy Environ. mechanisms must have the necessary capacity and Sci., , DOI:10.1039/C7EE02342A. injec vity, and must confi ne the CO and impede its 3 lateral migra on and/or ver cal leakage to other strata, D. Sandalow, R. Aines, J. S. Friedman, C. McCormick and S. T. McCoy, Carbon Dioxide U liza on (COU): ICEF shallow potable groundwater, soils and/or atmosphere. Roadmap 2.0, Innova on for Cool Earth Forum, Tokyo, Such geological media are mainly oil and gas reservoirs Japan, 2017. and deep saline aquifers that are found in sedimentary 4 basins. Storage of gases, including CO, in these media IEAGHG, Techno-Economic Evalua on of SMR Based Standalone (Merchant) Hydrogen Plant with CCS, IEA has been demonstrated on a commercial scale by Greenhouse Gas R&D Program, Cheltenham, UK, 2017. enhanced oil recovery opera ons, natural gas storage 5 and acid gas disposal. Some of the risks associated IEA, 20 Years of Carbon Capture and Storage: Accelera ng Future Deployment, Interna onal Energy with CO capture and geological storage are similar Agency, Paris, France, 2016. to, and comparable with, any other industrial ac vity for which extensive safety and regulatory frameworks 6 Enhance Energy Inc. and North West Redwater are in place. Specifi c risks associated with CO storage Partnership, Alberta Carbon Trunk Line project : knowledge sharing report, 2017, Calgary, Alberta, 2018. relate to the opera onal (injec on The prac cal global 7 Shell, Quest Carbon Capture and Storage Project: capacity to store CO2 in saline forma ons is believed, on the basis of geological assessments for select regions Annual Summary Report- Alberta Department of of the world, to be upwards of 4,000 GtCO . Another Energy: 2017, Shell Canada Energy, Calgary, Alberta, 2 2018. 1,000 GtCO2 are es mated to be available in depleted oil and gas reservoirs.17 While these capacity numbers 8 IEAGHG, The Carbon Capture Project at Air Products’ are large, they hide the uneven geological distribu on Port Arthur Hydrogen Produc on Facility, IEA Greenhouse Gas R&D Programme, Cheltenham, United of storage resources and the challenge of characterizing, Kingdom, 2018. assessing and developing storage sites, which can be high risk and take many years to a decade. This meline 9 Z. Liang, W. Rongwong, H. Liu, K. Fu, H. Gao, F. Cao, can also be complicated by the laws and regula ons R. Zhang, T. Sema, A. Henni, K. Sumon, D. Nath, D. Gelowitz, W. Srisang, C. Saiwan, A. Benamor, M. Al- that may—or may not—exist to facilitate safe, eff ec ve Marri, H. Shi, T. Supap, C. Chan, Q. Zhou, M. Abu-Zahra, 18 CO2 storage. 2005 Many governments have recognized M. Wilson, W. Olson, R. Idem and P. Ton wachwuthikul, that coordina ng and suppor ng development of CO2 Interna onal Journal of Greenhouse Gas Control, 2015, storage sites is as important to future deployment, if not 40, 26–54. more so, than support for technology development.5 10 R. Stanger, T. Wall, R. Spörl, M. Paneru, S. Grathwohl, Consequently, there is growing emphasis on M. Weidmann, G. Sche necht, D. McDonald, K. development of storage-centric and integrated CCUS Myöhänen, J. Ritvanen, S. Rahiala, T. Hyppänen, J. projects that are focused around clusters of emi ers Mletzko, A. Kather and S. Santos, Interna onal Journal (e.g., the Port of Ro erdam, Northern Alberta) and of Greenhouse Gas Control, 2015, 40, 55–125. linked to transport and storage hubs (e.g., in the North 11 J. C. Abanades, B. Arias, A. Lyngfelt, T. Ma sson, D. Sea, EOR in Southern Alberta).5,19 Many of these hub and E. Wiley, H. Li, M. T. Ho, E. Mangano and S. Brandani, cluster projects involve the cement, iron and steel, and Interna onal Journal of Greenhouse Gas Control, 2015, chemical industries. 40, 126–166. 12 K. [Na onal E. T. Lab. (NETL) Callahan Albany, OR 1 B. Metz, O. Davidson, H. de Coninck, M. Loos and L. (United States)], L. [Na onal E. T. Lab. (NETL) Goudarzi Meyer, Eds., IPCC Special Report on Carbon Dioxide Albany, OR (United States)], M. [Na onal E. T. Lab.

December 2019 23 (NETL) Wallace Albany, OR (United States)] and R. [Booz 16 S. Bachu, Progress in Energy and Combus on Science, A. H. Wallace McLean, VA (United States)], A Review 2008, 34, 254–273. of the CO Pipeline Infrastructure in the U.S., United 17 S. M. Benson, K. Bennaceur, P. Cook, J. Davison, H. de States, 2015. Coninck, K. Farhat, A. Ramirez, D. Simbeck, T. Surles, 13 P. C. Psarras, S. Comello, P. Bains, P. Charoensawadpong, P. Verma and I. Wright, in Global Energy Assessment S. Reichelstein and J. Wilcox, Environ. Sci. Technol., - Toward a Sustainable Future, Cambridge University 2017, 51, 11440–11449. Press, Cambridge, UK and New York, NY, USA and the Interna onal Ins tute for Applied Systems Analysis, 14 CSA, Geological storage of carbon dioxide, Canadian Laxenburg, Austria, 2012, pp. 993–1068. Standards Associa on, 2012. 18 T. Dixon, S. T. McCoy and I. Havercro , Interna onal 15 ISO, Carbon dioxide capture, transporta on and Journal of Greenhouse Gas Control, 2015, 40, 431–448. geological storage -- Geological storage, Interna onal Standards Organiza on, 2017. 19 GCCSI, Understanding industrial CCS hubs and clusters, Global CCS Ins tute, Melbourne, 2016.

December 2019 24 In 2014, produc on of cement contributed over 2 GtCO CHAPTER 3 to global greenhouse gas (GHG) emissions (about 6% of the global total).4,5 Of this amount, around 40% was emi ed from the use of fossil fuels to provide process CASE STUDIES heat in clinker produc on, while the remaining 60% was emi ed directly from the chemical decomposi on of CEMENT limestone (see Box 3-1). These fi gures do not include indirect emissions from genera on of electricity used in Industry Overview cement manufacturing (e.g., crushing and conveying of Cement is the founda on for the built environment. materials) or mining of limestone and other minerals. When combined with aggregates and water, cement makes the concrete used in roads, runways, buildings, Decarbonization pathways bridges, dams and other structures on which socie es Many strategies for decarbonizing cement produc on around the world depend. have been proposed. Some have been implemented. In recent years, over 4 Gt of cement have been produced There are many technical op ons for reducing the CO 3,6,7,8 annually.1,2 Global demand for cement has been growing footprint of cement produc on directly, as well as rapidly, expanding by nearly a factor of four between a growing literature that considers the problem more 1,9 1990 and 2014, when it reached around 600 kg/capita.1 holis cally. Approaches proposed to date include: The vast majority of growth in produc on since 1990 has ■ Switching to less carbon-intensive fuels, such as occurred in China, making China’s per capita produc on sustainable biomass and wastes that would otherwise today triple the global average. Global demand for be incinerated or improperly landfi lled—a prac ce cement is expected to con nue growing, with some that is becoming more common today.3,7,10 rebalancing of supply and demand at the regional level ■ Improving the effi ciency of exis ng cement plants (including reduced produc on in China and increased through retrofi ts that reduce both thermal energy produc on in India and other Asia-Pacifi c countries).3 demand in clinker produc on and overall electricity demand.3,7,10 BOX 31 Cement and Concrete Concrete that is used in the construc on industry is a mixture of cement (some mes referred to as binder), water and solid aggregates such as sand, gravel and crushed stone (some mes referred to as fi ller). A typical mixture by volume is 10-15% cement, 15-20% water and 60-75% aggregate. Manufacturing tradi onal Portland cement involves hea ng limestone (calcium carbonate) and other minerals (primarily aluminosilicates) in a kiln to form a material known as clinker, which is mixed with other cons tuents (e.g., gypsum, fl y ash, steel slag) and ground into a fi ne powder. When this is mixed with water and aggregates, a series of chemical processes (“curing”) converts the cement powder into interlocking crystals, which grow stronger over mme.e. These crystals give concrete very good compression strength—so it can support a lotot of weight—but poor tension strength, meaning that it cannot resist being pulled apart unless other materials are added, such as steel (“rebar”). The length of me it takes concrete to reach its required strength is referred to as “design age” and is impacted by the amount and formula on of the cement, as well as curing condi ons. The hydrated calcium oxide found in cement is very reac ve with CO and, in fact, the cement naturally absorbs CO out of the atmosphere over its life— although this is only a small frac on of that released during clinker produc on.

December 2019 25 ■ Waste heat recovery for electricity produc on that heat needed in cement making. This case study reviews reduces demand for higher carbon intensity off site the way heat is provided in manufacturing of OPC and power genera on.3,10 then examines the poten al benefi ts and costs of fuel ■ Applica on of carbon capture and storage (CCS) to subs tu on, CCS, hydrogen and electricity. reduce the emissions from both fuel combus on and decomposi on of limestone (calcina on) to lime The cement manufacturing process during clinker produc on.3,7,10–13 OPC is composed of calcium carbonate, clay and lesser ■ Reducing the clinker content of cement through the amounts of other minerals (e.g., sand, bauxite and addi on of supplementary carbonaceous materials alumina). These materials are crushed, mixed together (SCMs)—e.g., fl y ash, steelmaking slag, limestone and in specifi c propor ons and ground into a raw meal calcined clay—and op mizing the choice of SCMs for that is heated to produce Portland cement clinker. The the applica on.1,3,7,10,14,15 clinker is then mixed with rela vely small amounts of ■ Op mizing design of concrete structures and choice gypsum (calcium sulphate)—added to slow se ng of of concrete formula ons to use concrete more the cement—and ground into the fi ne powder that is effi ciently,1 and increasing the design age of concrete OPC. It is in these later steps that SCMs and fi llers (e.g., structures—i.e., allowing a longer me for the cement limestone) can be added to create specialized cements. to reach the design strength—in order to reduce the This process is illustrated inOPC is composed of calcium amount of cement required.14 carbonate, clay and lesser amounts of other minerals ■ Development and applica on of alterna ve binders (e.g., sand, bauxite and alumina). These materials for cement,1,10 such as beliteye’elimite-ferrite (BYF) are crushed, mixed together in specifi c propor ons clinkers,16 carbonate calcium silicate clinkers (CCSC),16 and ground into a raw meal that is heated to produce or alkali-ac vated materials (AAM).17 Portland cement clinker. The clinker is then mixed with ■ Industrializa on of cement produc on in emerging rela vely small amounts of gypsum (calcium sulphate)— economies to increase effi ciency of produc on added to slow se ng of the cement—and ground into processes, improve quality control and reduce overall the fi ne powder that is OPC. It is in these later steps that waste genera on.1,9 SCMs and fi llers (e.g., limestone) can be added to create ■ Be er managing concrete waste from demoli on in specialized cements. This process is illustrated inOPC is composed of calcium carbonate, clay and lesser amounts order to accelerate natural uptake of CO2 through carbona on of the ac ve phases in cement, which of other minerals (e.g., sand, bauxite and alumina). could theore cally result in full uptake of the original These materials are crushed, mixed together in specifi c emission from calcina on, albeit in the distant propor ons and ground into a raw meal that is heated future.18,19 to produce Portland cement clinker. The clinker is then mixed with rela vely small amounts of gypsum (calcium Some op ons in this menu (e.g., use of biomass fuels, sulphate)—added to slow se ng of the cement—and effi ciency improvements and waste heat recovery) ground into the fi ne powder that is OPC. It is in these are rela vely straigh orward, with li le impact on the later steps that SCMs and fi llers (e.g., limestone) can cement making process or resul ng product. Other be added to create specialized cements. This process is op ons (e.g., CCS) would require more substan al illustrated in Figure 3A-1, which represents today’s state- and capital-intensive modifi ca ons. Some op ons of-the-art cement-making process., which represents would require changes in the way cement is used (e.g., today’s state-of-the-art cement-making process., which op miza on of design, increased design age) or even represents today’s state-of-the-art cement-making replacement of conven onal cements to some extent process. (e.g., alterna ve binders). Of these op ons, the only In the state-of-the-art “dry-kiln” process (OPC is two that would directly result in reduc ons in process composed of calcium carbonate, clay and lesser amounts heat emissions from ordinary Portland cement (OPC) of other minerals (e.g., sand, bauxite and alumina). manufacturing are subs tu on of sustainable biomass These materials are crushed, mixed together in specifi c for fossil fuels and CCS. Li le has been wri en about propor ons and ground into a raw meal that is heated the use of hydrogen and electricity for provision of the to produce Portland cement clinker. The clinker is then

December 2019 26 Figure 3A-1. Current state-of-the-art dry-kiln Portland cement manufacturing process. mixed with rela vely small amounts of gypsum (calcium remainder occurs in the rotary kiln.10,13 The hot clinker sulphate)—added to slow se ng of the cement—and exi ng the rotary kiln is then cooled, prehea ng the air ground into the fi ne powder that is OPC. It is in these that is used in the kiln and pre-calciner. later steps that SCMs and fi llers (e.g., limestone) can As this descrip on implies, diff erent temperatures are be added to create specialized cements. This process required in the cement making process and the process is illustrated inOPC is composed of calcium carbonate, is highly heat integrated. This means that the op mal clay and lesser amounts of other minerals (e.g., sand, means of providing heat to the pre-calciner may not bauxite and alumina). These materials are crushed, be the same as in the kiln. In addi on, changes that mixed together in specifi c propor ons and ground into might impact the fl ow rates of gas in the system must a raw meal that is heated to produce Portland cement be carefully evaluated, as they may impact the clinker clinker. The clinker is then mixed with rela vely small capacity of the plant. Moreover, because combus on amounts of gypsum (calcium sulphate)—added to slow gases are in direct contact with the cement, the impact se ng of the cement—and ground into the fi ne powder of changes to the fuel composi on must be carefully that is OPC. It is in these later steps that SCMs and fi llers considered to avoid nega vely impac ng product quality. (e.g., limestone) can be added to create specialized cements. This process is illustrated in Figure 3A-1, which Substituting biofuels for fossil fuels represents today’s state-of-the-art cement-making process., which represents today’s state-of-the-art In 2014, coal provided 70% of the direct thermal energy cement-making process.), raw meal is conveyed to input to cement manufacturing globally, followed by 3 the top of a tower that holds a series of 3-6 cyclone oil and gas at 24% and “alterna ve” fuels at 6%. The separators arranged one above another. As the meal alterna ve fuel category includes both waste (e.g., descends through this series of cyclones, it is gradually municipal solid waste, sewage sludge and hazardous heated by contac ng hot exhaust gases, which are, in wastes) and bioenergy crops, although waste dominates a turn, gradually cooled. These hot exhaust gases come in energy terms . Where coal is used as the fuel input to from burning fuel in the calciner and rotary kiln. Meal the state-of-the-art described earlier, direct emissions is directly heated to around 900 °C by burning fuel in from coal combus on contribute around 300 kgCO2-e of the calciner, which is located near the end of the series the total 900 kgCO2-e of direct and indirect (electrical) of cyclones. Meal passes from there into the rotary emissions for each metric ton of clinker; where gas is kiln, where the meal is fi nally converted into nodules used, emissions from fuel combus on fall to 180 kgCO2- of clinker at temperatures of around 1,450 °C (and gas temperatures of nearly 2,000 °C). Of the total heat input a High temperatures and long residence mes in the cement making to the plant, 60-70% occurs in the pre-calciner and the process make it eff ec ve for incinera on of waste (and hazardous waste in par cular).

December 2019 27 Figure 3A-2. The greenhouse gas footprint of clinker produced using alterna ve fuels, coal with CCS op ons, hydrogen and electricity. These results include combus on emissions from process heat and calcina on in the cement plant, genera on of electricity, and produc on of hydrogen (but not upstream emissions from fossil fuels produc on). The coal baseline and CCS op ons are based on data presented in the CEMCAP project.22

e/t clinker as shown in Figure 3A-2b. Switching from equals the amount of CO sequestered in biomass. This fossil fuels to waste materials can result in emissions conven on, widely adopted in life cycle assessment (LCA reduc ons that vary as a func on of the source from Further, upstream emissions associated with agriculture which the waste material is derived and how it would (e.g., fer lizers, harves ng and transport) and land-use have otherwise been disposed. For example, using change are non-zero and should be a ributed to the biomass-based wastes (e.g., wood or agricultural fuel.24 Waste biomass would, by conven on, not carry processing wastes) would generally result in the direct these emissions burdens. Thus, on a lifecycle basis, emissions from combus on being reduced to zero as, use of bioenergy crops for process heat would result in

by conven on, the CO2 emi ed during combus on was lesser emissions reduc ons than biomass-based wastes. drawn from the atmosphere during growth (as assumed There is a limited sustainable biomass (both waste and in Figure 3A-2). As a result, in the 2017 IEA 2DS scenario, crop-based) available for use in materials or as fuels the share of waste (and biomass fuels) used directly in globally, so there will likely be strong compe on for cement manufacturing grows strongly by 2050.5 biomass that will limit its cost-eff ec veness in prac ce.3 Emissions reduc ons that might emerge from use Because biomass-based fuels tend to have a lower of bioenergy crops (e.g., switchgrass, poplar) are energy content (per unit mass) than fossil fuels—or, as somewhat more complicated to evaluate. (See Chapter will be discussed, hydrogen—they are not suitable for 2B.) While the direct emissions would also be zero providing the high temperatures required in the kiln by conven on, in reality there is a ming diff erence directly but can be used in the calciner to provide lower

between CO2 emissions and uptake by biomass that can temperature heat. For example, bioenergy crops such as have signifi cant impact for long rota on crops.23i.e. the miscanthus, switchgrass, poplar and pine all have a lower CO released from biofuel combus on approximately hea ng value of 17-19 GJ/t25, whereas most kilns require a fuel with a hea ng value of 20-22 GJ/t at a minimum.10 Burning alterna ve fuels in the calciner requires use b Assuming an electric grid intensity of 519 gCO/kWh,5 IPCC emissions factors20 and other energy requirements, as detailed in the CEMCAP of mul -channel burners in the calciner and careful base case.13 These fi gures do not include the upstream emissions from monitoring of the levels of impuri es, such as chloride, produc on and distribu on of fossil fuel or mining of raw materials. In the case of natural gas, for example, inclusion of these emissions would in the clinker.10 However, these issues are handled in increase emissions by around 20% per ton of OPC based on recent prac ce today at facili es that burn alterna ve fuels and es mates of average US natural gas upstream emissions.21

December 2019 28 do not appear to be a major barrier. A higher degree of CaCO3—the same material found in raw meal for OPC—

subs tu on of biomass and waste for other alterna ve and the CO2 is then driven off by the same calcina on fuels could be achieved by gasifi ca on of the feedstocks process. The CaL process can be added to the cement and subsequent use of the syngas (i.e., a mixture of plant in either the “tail-end” confi gura on, which carbon monoxide and hydrogen) as fuel, but this is at the requires no major modifi ca ons to the cement making research stage in the cement sector.11 process, or in the “integrated” confi gura on by using a shared calciner, which entails major modifi ca ons to the Application of CCS to cement making cement plant. The tail-end confi gura on has no impact The poten al importance of CCS to reduce emissions on cement produc on, while the integrated process from cement manufacturing was recognized by the late- could impact cement quality. In both cases, fuel and 1990s26 and has been emphasized through successive limestone consump on increases and, depending on analyses and road-mapping ac vi es.3,7,10,11,27 The value the confi gura on and design choices, the plant could 13 of CCS to the cement industry is that it could reduce become a net generator of electricity. the chilled ammonia process, membrane-assisted CO liquefac on, direct emissions from cement manufacturing—both 2 from process heat and calcina on—by 95%13 the chilled and the calcium looping process with tail-end and ammonia process, membrane-assisted CO liquefac on, integrated confi gura ons. For comparison, absorp on and the calcium looping process with tail-end and with monoethanolamine (MEA The CaL process has integrated confi gura ons. For comparison, absorp on been demonstrated at MW-equivalent scale for post- 28,29 with monoethanolamine (MEA and, when coupled combus on capture in power genera on, is being 10 with alterna ve fuels, result in zero (or even nega ve) pilot-tested by Taiwan Cement Corpora on and will be c emissions. Thus, CCS has an important role to play further tested in the EU-funded Cleanker Project in Italy. amongst the mul ple emissions mi ga on op ons for Oxyfuel technology could also be applied to cement the cement industry. plants by conver ng the precalciner and, op onally, the While all three classes of CO capture technology kiln to use pure oxygen (rather than air) for combus on 2 of the fuel and by recycling some of the CO -rich fl ue gas (see Chapter 2D) could be applicable to cement 2 11,30 manufacturing, post-combus on and oxy-combus on to control temperature in combus on. This would technologies are seen as the leading candidates.10 necessitate changes to the burners in the precalciner Solvent-based, post-combus on capture technologies and kiln, modifi ca ons to reduce air leakage into the (e.g., amine-based solvents, chilled ammonia) can system and changes to the clinker cooler, but it requires be added to a cement plant without making major less signifi cant modifi ca ons than the integrated CaL 13,30 modifi ca ons to the cement-making process or confi gura on. It would require addi onal electricity impac ng cement produc on. However, both require (for air separa on), but no addi onal fuel input. The increased concentra on of CO in the kiln, precalciner substan al amounts of addi onal steam that would 2 need to be generated on site (from addi onal fuel) and preheater would impact heat transfer and the or imported from neighboring facili es.13the chilled degree of calcina on of the product (at a constant ammonia process, membrane-assisted CO liquefac on, temperature), which could impact product quality if not 30 and the calcium looping process with tail-end and appropriately managed. Oxyfuel technology has been integrated confi gura ons. For comparison, absorp on pilot-tested at a Lafarge cement plant in 2011 and 2012 11,30 with monoethanolamine (MEA Amine-based capture in Denmark and has been widely studied. systems have been the focus of many engineering Applying CCS to cement produc on would reduce the studies, and one was successfully pilot-tested by Norcem combined direct process heat and calcina on emissions between 2013 and 2017 at a plant in Norway.10,13 from cement making by upwards of 95%. However, In addi on to the conven onal solvent-based op ons, the overall reduc on depends on the type of capture calcium looping (CaL) technology has also been system applied and the carbon intensity of energy 13 inves gated for applica on to cement plants for post- inputs (e.g., supplemental fuel, electricity). the chilled ammonia process, membrane-assisted CO liquefac on, combus on capture. CaL shares many similari es with 2

cement making, as it is a cyclic process in which CO2 is captured from fl ue gas using CaO, which results in c See h p://www.cleanker.eu/home-page-it.html

December 2019 29 and the calcium looping process with tail-end and Conven onal approaches to CO2 emission reduc on in integrated confi gura ons. For comparison, absorp on cement plants are based on post-combus on capture with monoethanolamine (MEA For example, in the case with chemical solvents, requiring a substan al energy of a cement plant that uses coal as the primary fuel consump on for regenera on, or oxycombus on in and natural gas to provide steam for an amine-based the cement kiln, involving a deep redesign of the plant. capture system (without capture) and draws electricity The aim of this work is inves ga ng the applica on of

at the global average emissions intensity, the overall Molten Carbonate Fuel Cells in cement plants for CO2 emissions reduc on would be closer to 60% (i.e., from capture from the plant exhaust gases, using the fuel

900 to 370 kgCO2-e/t clinker) as shown in Figure 3A-2. cells as ac ve CO2 concentrators of combus on fl ue On the other hand, the integrated CaL process would gases and eventually obtaining a purifi ed CO stream reduce emissions by upwards of 80% from a coal-fueled through a cryogenic process. A novel confi gura on with

baseline to around 140 kgCO2-e/t clinker. Unfortunately, MCFCs added along the exhaust line has been assessed these substan al emission reduc ons would also result by means of process simula ons. The results show a

in an increase of 50% (MEA) to 80% (integrated CaL) in remarkable poten al in terms of equivalent avoided CO2 the cost of clinker produc on, rela ve to a state-of-the- emissions (exceeding 1000g/kWh may also prove to be art baseline. This corresponds to mi ga on costs of an a rac ve op on when integrated into the cement 12 $100 and $70/tCO2 avoided, respec vely. making process.

None of these CO2 capture technologies have yet been demonstrated at full-scale on a cement plant. However, Use of hydrogen as an alternative fuel this may soon change, as the Norwegian government is Hydrogen is not currently used as an alterna ve inves ng in development of an integrated CCS project fuel in cement making, nor has use of hydrogen as

that would capture CO2 from the Norcem cement plant an alterna ve fuel in cement making been widely at Brevik, Norwayd. If funded, this demonstra on project inves gated. While the combus on temperature of

would not only provide valuable informa on about CO2 hydrogen in air is more than suffi cient to provide capture in the cement industry, but would also develop the temperatures required in the cement kiln, the

the CO2 transport and storage infrastructure that is combus on proper es of hydrogen mean that it currently missing in Europe. The absence of transport would need to be burnt in specially designed burners and storage infrastructure poses a barrier to deployment or mixed with solid par cles (e.g., clinker dust) to be of CCS across industries globally and will need to be eff ec ve.33 How use of hydrogen impacts clinker quality addressed if this emissions reduc on op on is to be also appears to be an open ques on. Given the state of taken up for cement or any other sector. research on hydrogen for cement making, there may be other unknown issues as well.33 Given that the most promising technologies available for capture from cement produc on are diff erent Technical challenges of hydrogen combus on from those in power genera on (e.g., CaL) and, for notwithstanding, use of green hydrogen generated maximum emissions reduc on benefi t, should be from electrolysis with fully renewable electricity would integrated into the process, con nued focused research eliminate process heat emissions as shown in Figure and pilot tes ng would be benefi cial. For example, the 3A-2—an overall reduc on of around 30% rela ve to LEILAC project (an EU-funded research project) aims a coal-fueled plant. Of course, the intermi ent nature to demonstrate Direct Separa on calcining technology of renewables would necessitate energy (or hydrogen for cement manufacture at pilot-scale, in which raw storage) to support con nuous opera on of a cement meal is heated indirectly, separa ng fuel combus on plant, so alterna ves would be to use grid electricity from the calcina on process.31 Other emerging for electrolysis or grey hydrogen from steam-methane capture technologies, such as molten carbonate reforming (SMR). At the current global average grid fuel cells,32necessary for sustaining the endothermic intensity, however, the emissions from a plant that uses calcina on process and the forma on of clinker. hydrogen generated using grid electricity would be 1.4 mes that of a coal-based plant (Figure 3A-2). The grid

emissions factor would need to be around 250 kgCO2/ d See h ps://www.norcem.no/en/carbon_capture

December 2019 30 kWh to achieve the same overall emissions intensity as the capacity of commercially available rotary calciners

a coal-fi red cement plant and around 180 kgCO2/kWh to are many mes smaller than those used in state-of- meet that of an SMR plant (without CCS). Extending the the-art cement plants and maximum temperatures are analysis illustrated in Figure 3A-2 to include upstream somewhat lower (as discussed in Sec on 2C), sugges ng emissions associated with natural gas supply (and further development is requirede. renewable genera on) would likely favor blue hydrogen For the same energy input, the cost of using electricity (i.e., from SMR with CCS) over electrical routes. This directly to generate heat (at nearly 100% effi ciency) highlights the interlinked nature of hydrogen with would be lower than using hydrogen produced via decarboniza on of the electricity sector. electrolysis (as today, electrolysis at large scales is As in the case of biofuel subs tu on, use of hydrogen around 70% effi cient).34 At the same me, in a system for process heat has no impact on the emissions from that combined Direct Separa on calcining technology calcina on or other indirect emissions from electricity with an electrically heated rotary calciner, all of the genera on. Thus, subs tu on of hydrogen would be calcina on emissions could be captured. This could limited to a 30% reduc on rela ve to a coal-fi red case, result in lower overall avoidance cost than the hydrogen all else being equal. This would increase the overall cost routes and could poten ally be compe ve with current of clinker produc on by at least a factor of three (for approaches to CCS. However, considerable further green hydrogen) on the basis of fuel alone, making the research is needed to assess these technologies before overall cost of emissions mi ga on much higher than any strong conclusions can be drawn. for direct applica on of CCS. Should Direct Separa on calcina on technology be shown to be viable,31 it Conclusion could make hydrogen subs tu on a more a rac ve Manufacturing of cement is a substan al contributor op on. Such a combina on would allow capture of the to global GHG emissions. As the demand for cement is calcina on emissions, somewhat reducing the mi ga on projected to grow in the decades ahead, the sector will cost. Nonetheless, given the rela vely low poten al need to aggressively reduce its emissions to reach levels and high cost of emissions reduc ons through use of consistent with a 2 °C or 1.5 °C target. hydrogen fuel, it may not be a valuable approach to pursue in a deep decarboniza on scenario. A wide range of op ons for emissions reduc ons have been evaluated, ranging from effi ciency improvements Direct electrical heating to subs tu on of SCMs for emissions-intensive clinker to changes in the way cement is used in construc on. Li le has been wri en about direct electrical hea ng Around 40% of the direct CO emissions from cement for cement manufacturing. It should be technically 2 manufacturing are associated with process heat for possible to heat raw meal to a suffi cient temperature to clinker produc on, while the remainder are generated decompose CaCO and form the ac ve phases of cement 3 from decomposi on of calcium carbonate in the process. using electrical-resistance hea ng (or other methods Given that the vast majority of cement is produced discussed in Sec on 2C). As in the hydrogen case, if using coal for heat today, subs tu on of lower-carbon Direct Separa on calcining technology is successful, it intensity fuels is already having a substan al impact. would not only allow CO resul ng from calcina on of 2 Biomass-based wastes and sustainable biofuels have the raw meal to be captured in a concentrated form and an important role to play, but they can only subs tute geologically stored but would also enable effi cient use of for a frac on of the heat input in cement making (due other heat sources such as electrical-resistance hea ng. to their low hea ng value) and, given their limited However, Direct Separa on calcining (using electricity, supply, may not be cost eff ec ve in large quan es. or otherwise) would only allow for a par al replacement Reducing emissions from hea ng via CCS is an important of process heat with electricity, as it does not replace op on that has been researched for many years and the rotary kiln. Indirectly heated rotary calciners—i.e., is now being demonstrated. While rela vely costly a rotary kiln to which heat is provided by natural gas or in comparison to use of alterna ve fuels (and other electric resistance elements—are manufactured today for specifi c applica ons where the atmosphere in the calciner must be controlled, such as pyrolysis. However, e For example, indirectly heated kilns are available from IBU-tec, FEECO, and Kurimoto Ltd.

December 2019 31 mi ga on op ons), it has the dis nct benefi t of being 5 IEA, Energy Technology Perspec ves 2017: Catalysing able to reduce the total emissions—both process heat Energy Technology Transforma ons, Interna onal and calcina on—by upwards of 90%. Use of hydrogen, Energy Agency, Paris, France, 2017. green or otherwise, as an alterna ve fuel appears 6 D. Sandalow, R. Aines, J. S. Friedman, C. McCormick to have less poten al, as it is more costly than direct and S. T. McCoy, Carbon Dioxide U liza on (COU): ICEF applica on of CCS in the cement plant and would reduce Roadmap 2.0, Innova on for Cool Earth Forum, Tokyo, only the emissions from process heat. Direct Separa on Japan, 2017. technology, however, could enable more cost-eff ec ve 7 IEA, Cement Technology Roadmap 2009: Carbon emissions reduc ons from hydrogen subs tu on, direct emissions reduc ons up to 2050, Interna onal Energy electrical hea ng and carbon capture. There may also Agency, Paris, France, 2009. be op ons that have not yet been iden fi ed in literature 8 E. Benhelal, G. Zahedi, E. Shamsaei and A. Bahadori, to combine fuel subs tu on with CCS to achieve zero or Journal of Cleaner Produc on, 2013, 51, 142–161. even nega ve emission cement. 9 S. A. Miller, V. M. John, S. A. Pacca and A. Horvath, This case study highlights the challenges of trying to Cement and Concrete Research, 2018, 114, 115–124. separately address process heat emissions from those 10 ECRA and CSI, Development of State of the Art- resul ng from carbon in the feedstock and highlights Techniques in Cement Manufacturing: Trying to Look the need to examine emission-reduc on op ons in Ahead, Revision 2017, European Cement Research industrial processes in an integrated fashion. Con nued Academy, Duesseldorf, Germany, 2017. research, development and demonstra on in the area 11 IEAGHG, Deployment of CCS in the Cement Industry, of emissions reduc ons in cement-making is needed to IEA Greenhouse Gas R&D Programme, Cheltenham, achieve zero emissions processes. United Kingdom, 2013. 12 S. O. Gardarsdo r, E. De Lena, M. Romano, S. 1 K. L. Scrivener, V. M. John and E. M. Gartner, Eco- Roussanaly, M. Voldsund, J.-F. Pérez-Calvo, D. Berstad, effi cient cements: Poten al, economically viable C. Fu, R. Anantharaman, D. Su er, M. Gazzani, M. solu ons for a low-CO, cement based materials Mazzo and G. Cin , Energies, 2019, 12, 542. industry, United Na ons Environment Programme, 13 M. Voldsund, S. O. Gardarsdo r, E. De Lena, J.-F. Paris, France, 2016. Pérez-Calvo, A. Jamali, D. Berstad, C. Fu, M. Romano, S. 2 R. M. Andrew, Earth Syst. Sci. Data, 2018, 10, 195–217. Roussanaly, R. Anantharaman, H. Hoppe, D. Su er, M.

3 Mazzo , M. Gazzani, G. Cin and K. Jordal, Energies, IEA, Technology Roadmap: Low-Carbon Transi on in the 2019, 12, 559. Cement Industry, Interna onal Energy Agency, Paris, France, 2018. 14 S. A. Miller, A. Horvath and P. J. M. Monteiro, Environ. Res. Le ., 2016, 11, 074029. 4 C. Le Quéré, R. M. Andrew, P. Friedlingstein, S. Sitch, J. Hauck, J. Pongratz, P. A. Pickers, J. I. Korsbakken, 15 K. Scrivener, F. Mar rena, S. Bishnoi and S. Maity, G. P. Peters, J. G. Canadell, A. Arneth, V. K. Arora, L. Cement and Concrete Research, 2018, 114, 49–56. Barbero, A. Bastos, L. Bopp, F. Chevallier, L. P. Chini, P. 16 E. Gartner and T. Sui, Cement and Concrete Research, Ciais, S. C. Doney, T. Gkritzalis, D. S. Goll, I. Harris, V. 2018, 114, 27–39. Haverd, F. M. Hoff man, M. Hoppema, R. A. Houghton, G. Hur , T. Ilyina, A. K. Jain, T. Johannessen, C. D. Jones, 17 J. L. Provis, Cement and Concrete Research, 2018, 114, E. Kato, R. F. Keeling, K. K. Goldewijk, P. Landschützer, 40–48. N. Lefèvre, S. Lienert, Z. Liu, D. Lombardozzi, N. Metzl, 18 F. Xi, S. J. Davis, P. Ciais, D. Crawford-Brown, D. Guan, C. D. R. Munro, J. E. M. S. Nabel, S. Nakaoka, C. Neill, A. Pade, T. Shi, M. Syddall, J. Lv, L. Ji, L. Bing, J. Wang, W. Olsen, T. Ono, P. Patra, A. Peregon, W. Peters, P. Peylin, Wei, K.-H. Yang, B. Lagerblad, I. Galan, C. Andrade, Y. B. Pfeil, D. Pierrot, B. Poulter, G. Rehder, L. Resplandy, Zhang and Z. Liu, Nature Geosci, 2016, 9, 880–883. E. Robertson, M. Rocher, C. Rödenbeck, U. Schuster, J. Schwinger, R. Séférian, I. Skjelvan, T. Steinhoff , A. 19 C. Pade and M. Guimaraes, Cement and Concrete Su on, P. P. Tans, H. Tian, B. Tilbrook, F. N. Tubiello, Research, 2007, 37, 1348–1356.

I. T. van der Laan-Luijkx, G. R. van der Werf, N. Viovy, 20 A. P. Walker, A. J. Wiltshire, R. Wright, S. Zaehle and H. S. Eggleston, L. Buendia, K. Miwa, T. Ngara and B. Zheng, Earth System Science Data, 2018, 10, K. Tanabe, Eds., 2006 IPCC Guidelines for Na onal 2141–2194. Greenhouse Gas Inventories, Ins tute for Global

December 2019 32 Environmental Strategies, Japan, 2006, vol. Volume 2: 28 B. Arias, M. E. Diego, J. C. Abanades, M. Lorenzo, L. Energy. Diaz, D. Mar nez, J. Alvarez and A. Sánchez-Biezma, Interna onal Journal of Greenhouse Gas Control, 2013, 21 J. A. Li lefi eld, J. Marrio , G. A. Schivley and T. J. Skone, 18, 237–245. Journal of Cleaner Produc on, 2017, 148, 118–126. 29 J. Hilz, M. Helbig, M. Haaf, A. Daikeler, J. Ströhle and B. 22 M. Voldsund, R. Anantharaman, D. Berstad, E. De Lena, Epple, Fuel, 2017, 210, 892–899. C. Fu, S. O. Gardarsdo r, A. Jamali, J.-F. Pérez-Calvo, M. Romano, S. Roussanaly, J. Ruppert, O. Stallmann and D. 30 F. Carrasco-Maldonado, R. Spörl, K. Fleiger, V. Hoenig, Su er, CEMCAP compara ve techno-economic analysis J. Maier and G. Sche necht, Interna onal Journal of of CO capture in cement plants (D4.6), 2019. Greenhouse Gas Control, 2016, 45, 189–199. 23 F. Cherubini, G. P. Peters, T. Berntsen, A. H. Strømman 31 T. P. Hills, M. Sceats, D. Rennie and P. Fennell, Energy and E. Hertwich, GCB Bioenergy, 2011, 3, 413–426. Procedia, 2017, 114, 6166–6170. 24 R. J. Plevin, J. Beckman, A. A. Golub, J. Witcover and M. 32 M. Spinelli, M. C. Romano, S. Consonni, S. Campanari, O’Hare, Environ. Sci. Technol., 2015, 49, 2656–2664. M. Marchi and G. Cin , Energy Procedia, 2014, 63, 6517–6526. 25 ECN.TNO, Phyllis2 - Database for biomass and waste, h ps://phyllis.nl/, (accessed August 16, 2019). 33 V. Hoenig, H. Hoppe and B. Emberger, Carbon Capture Technology - Op ons and Poten als for the Cement 26 IEAGHG, The reduc on of greenhouse gas emissions Industry, European Cement Research Academy, from the cement industry, IEA Greenhouse Gas R&D Duesseldorf, Germany, 2007. Programme, Cheltenham, UK, 1999. 34 IEA, Technology Roadmap: Hydrogen and Fuel Cells, 27 IEAGHG, CO Capture in the Cement Industry, IEA Interna onal Energy Agency, Paris, France, 2015. Greenhouse Gas R&D Programme, Cheltenham, UK, 2008.

December 2019 33 IRON AND STEEL averages 200 kg per capita globally but varies from a low of 30 kg per capita in Africa to a high of 283 kg per capita Industry overview in NAFTA countries. Since trade is global, with exports The global iron and steel industry is one of the largest in equal to 27% of produc on in 2018, prices are strongly the world, with sales of $2.5 trillion in 2017.1 The 2018 aff ected by global compe on.2 produc on of crude steel was 1,808 million tons, up 7% from 2017 and a 10-fold increase since 1950.2,3 Steel Decarbonization pathways is one of the largest products by weight produced by The two main processes used to produce steel are the humanity—one of very few commodi es manufactured older blast furnace/basic oxygen furnace route (BOF) at the gigaton scale—and it is used in a vast range of and the newer electric arc furnace route (EAF) (see industries including construc on, automo ve, shipping, Figure 3B-1). The BOF route begins with raw iron ore aerospace and energy equipment. Because of this and includes many processing steps. For this reason, it enormous scale and the fact that conven onal iron and is usually performed at rela vely large integrated steel steel produc on is energy- and emissions-intensive, mills that also incorporate facili es for sintering iron ore the sector is responsible for approximately 7% of and producing coke from coal. The EAF route primarily 4 global CO2 emissions. Finding prac cal approaches to uses recycled steel scrap as its feedstock and includes decarbonizing iron and steel produc on is therefore of fewer processing steps. It is therefore usually performed vital importance to achieving climate goals. at smaller, “mini-mill” facili es where steel scrap is Any successful approach to decarbonizing iron and steel widely available. must take into account the fact that the industry is in The dominant energy requirement in producing virgin a state of fl ux. A decade of rapid expansion and recent steel is extrac ng metallic iron from raw iron ore demand satura on has led to severe overcapacity, with (smel ng). In the BOF route, this is performed at blast average plant use at approximately 75%. If all planned furnaces, which consequently have very high emissions. steel produc on projects are realized, global capacity Steel that has already been produced and recycled as could increase by 4-5% between 2019 and 2021, pu ng scrap can be reprocessed via the EAF route without this further pressure on use.5,6 Steel produc on has also step, leading to very large energy and emissions savings been increasingly concentrated, with 51% of 2018 (as much as 90% reduc on8). Addi onally, since EAF is produc on in China (up from 38% in 2008), while India, almost en rely electrifi ed, powering it with low-carbon Japan, the US, Korea and Russia collec vely account electricity can almost en rely eliminate its emissions. for an addi onal 25%. Meanwhile, steel consump on

Figure 3B-1. Primary iron and steelmaking routes (adapted from Carpenter et al7). While both BOF and EAF routes can use hot metal, scrap and DRI as feedstocks, BOF primarily uses hot metal while EAF primarily uses scrap.

December 2019 34 As a result, the transi on from BOF to EAF results in signifi cantly reduced emissions. Globally, BOF accounts for approximately three-quarters of steel produc on and EAF for one quarter, down from its peak of 33% in 2000. The balance between BOF and EAF varies drama cally within diff erent countries, mostly as a result of the availability of scrap steel and reliable, low-cost electricity (see Figure 3B-2). For example, due to the long history of steelmaking in the US and Mexico and the large amount that is recycled, two-thirds of steel produc on is via EAF. The average emissions intensity of steel produc on is therefore

rela vely low at 1,736 kg CO2/ton (US) and 1,080 kg 3,9 CO2/ton (Mexico). (The local electric grid emissions intensity also impacts these values.) In China over 90% of steel produc on is via BOF, and the average emissions intensity is correspondingly higher at 2,148 9,10 kg CO2/ton crude steel. In considering the possible future adop on of EAF (displacing BOF), it is instruc ve to note that the Figure 3B-2. Share of steel produc on via EAF in diff erent region/ stock of steel per capita has historically saturated at countries; data from worldsteel.14,15 11-15 tons in countries that have fully industrialized.11 However, understanding the stock of steel and the iron suitable for the EAF route, but it avoids the use of amount available for recycling is complex and would coke, signifi cantly reducing its emissions. Expanding 12 benefi t from addi onal research. A scenario analysis of the use of DRI for EAF steel produc on is therefore a steel produc on based on mass fl ow analysis suggests poten al route to low-carbon virgin steel produc on. that the last required blast furnace for primary steel (See below for further discussion of DRI.) produc on could be built as early as 2020.13 While the EAF route primarily uses recycled steel scrap, Process heat in iron and steel making it is able to produce virgin steel using direct reduced iron It is important to note that while both BOF and EAF (DRI) as a feedstock. DRI processes iron ore to a metallic require large amounts of process heat, the two routes BOX 32 BOF vs EAF in China Because EAF steel produc on is so much less emissions-intensive, transi oning from BOF to EAF can lead to very large emissions reduc ons. However, doing so requires signifi cant capital investmentsnts and confi dence that there will be a suffi cient supply of recycled scrap steel available.e. This is of par cular interest in China, where steelmaking is dominated by BOF (see Figure 3B-2) but where there are signs of a growing availability of scrap steel in the medium-term. This issue has been underscored by a recent surge in available scrap steel due to the shu ering of ineffi cient induc on furnaces. However, many Chinese BOF facili es (par cularly blast furnaces) are rela vely new and effi cient, making their replacement less economically a rac ve. The evolu on of BOF vs EAF steelmaking in China will be an important driver of global industrial emissions over the next several decades.5,16

December 2019 35 Figure 3B-3. Blast furnace for ironmaking. Iron ore, coke and limestone are added in layers at the top, sinking slowly to the bo om. Hot air is blasted into the furnace, igni ng the coke in the combus on zone and producing CO gas and heat. Coke also physically supports the descending layers of iron ore and provides porosity for movement of CO gas and liquid iron. generate it in very diff erent ways. In BOF steelmaking, Process description: BOF process heat is provided through the combus on of BOF steelmaking is used to produce new steel from coke. Coke also serves three other func ons: chemical raw iron ore and other ingredients. The fi rst step in the reduc on of iron ore, physical support of the burden process is to combine iron ore—primarily magne te in blast furnaces, and porosity for hot gas and molten (Fe O ) and hema te (Fe O )—with limestone and fi ne metal movement (see below). This mul -purpose role 3 4 2 3 par cles of coke (coke breeze) in a process known as of coke means that simply replacing coke combus on sintering. The coke is ignited and burns at 1,300-1,480 with other sources of process heat is imprac cal. In °C, partly mel ng the materials and producing a coarse EAF steelmaking, process heat is provided electrically, agglomerate called sinter that is appropriate for bulk which also makes direct replacement unappealing as an handling. The heated sinter is usually cooled in open air emissions-reduc on strategy and also unnecessary to or with water sprays; waste heat recovery is technically the extent that the electricity comes from low-carbon feasible but rare.17,18 sources. The complexi es of process heat in iron and steelmaking therefore generally require signifi cant In parallel to sinter produc on, metallurgical-grade process change to decarbonize. coal is heated in the absence of air at 900-1,100 °C to produce coke. The hea ng drives off vola les and While EAF off ers many advantages, BOF steelmaking will leaves a nearly pure carbon mass that is both porous con nue to be a large share of global produc on. There and structurally strong. Heat is provided by combus ng is also a large installed base of blast furnaces and related a por on of the coke oven gas (COG), the remainder of equipment that represents a large investment of capital. which is usually used in other combus on units at an Finding technical methods to reduce emissions intensity integrated steel plant. At an integrated steel mill, both of the BOF route without major process changes is sintering and coking facili es are usually co-located with therefore of great interest and will form the bulk of the the rest of the iron and steel making facili es because of discussion to follow.

December 2019 36 their importance to the iron and steel making.9,19 providing heat and reducing the total carbon content. The resul ng molten steel and slag are then tapped at The sinter, coke, addi onal pellet or lump ore, and periodic intervals. In contrast to the blast furnace, the limestone are then fed to a large blast furnace for the steelmaking furnace requires no addi onal heat input iron smel ng process (see Figure 3B-3). The furnace because all of the necessary heat is provided by the is charged with materials from the top, which form oxygen-carbon reac ons. alterna ng layers and sink slowly towards the bo om. Water-cooled nozzles (tuyeres) inject air heated to 900-1,300 °C near the bo om. This blast air is heated in Emissions reduction in the BOF route

hot blast stoves, which primarily burn top gas recycled The CO2 emissions from BOF come from a wide variety from the blast furnace, possibly with the addi on of sources (see Figure 3B-4). The largest share (39%) is of some coke oven gas.20 This hot blast air burns the from fl ue gas emi ed from power genera on. However, coke, forming carbon monoxide (CO) and hea ng this this genera on is based on combus ng blast furnace combus on zone to 1,500 °C or higher. Fuel combusted gas (BFG) and coke oven gas (COG), complica ng the for hea ng the hot blast stoves results in the majority of a ribu on of these emissions (some natural gas may be 7 CO2 emi ed from a blast furnace. combusted as a supplementary fuel). The next largest share is from the stoves that heat air for injec on via The hot CO rises through the furnace and reacts with the tuyeres in the blast furnace (18%), with important the sinking iron ore, chemically reducing it to iron and contribu ons from the sinter plant and coke plant (both CO . The heat also decomposes the limestone into lime 2 16%).7 However, the hot air stoves, coke ovens and sinter (CaO) and CO ; the lime in turn reacts with trace silicon 2 plant also combust large por ons of BFG and/or COG, impuri es to form calcium silicate (CaSiO ) slag. The 3 illustra ng the close process integra on of the overall molten iron and slag sink to the bo om of the furnace BOF process. where they are tapped, while the CO2 and other hot gases (known collec vely as “blast furnace gas”) rise to While these are all process heat contribu ons to the top where they are used for a variety of purposes, emissions, replacing any one of them with an alterna ve including pre-hea ng the tuyere blast air and genera ng process hea ng method, such as combus ng hydrogen electricity. or providing electric heat, would disrupt this integra on. This would lead to a need to handle excess BFG and A key feature of the blast furnace is the alterna ng COG that is no longer being combusted, or (in the case addi on of charge materials to the top, which form layers that slowly sink. They are partly supported on the coke, which is structurally strong and slows their descent, while also being porous enough to allow molten metal to fl ow downwards and hot gases to fl ow upwards. This allows for a long-dura on interac on within the furnace (counter-current fl ow), op mizing the chemical reac ons. The iron that is tapped from the bo om of the blast furnace is called pig iron or hot metal. It is mostly pure iron, with a rela vely high carbon content (approximately 4%) compared to fi nished steel. At integrated steel mills, it is transferred directly into a steel-making furnace, along with up to approximately 30% scrap steel. Within the steelmaking furnace, a water- cooled lance injects pure oxygen, which reacts Figure 3B-4. CO2 emissions from diff erent processes within BOF with the remaining carbon in the hot metal, steelmaking. Adapted from Carpenter, 2012.7

December 2019 37 of coke oven modifi ca ons) reduce the amount of COG Steelmaking Process by Innova ve Technology for available for use by various processes. This integra on Cool Earth 50 (COURSE50) program in Japan has makes piecemeal process heat modifi ca on una rac ve experimented with the use of hydrogen reduc on for BOF technology and leads instead to an examina on in blast furnaces to reduce coke consump on and of carbon capture (see below) or op ons that favor emissions. Using coke oven gas, either directly or material subs tu on in the process, which include: reformed to increase the hydrogen content, the program demonstrated hydrogen reduc on at a test ■ Reduction or elimination of sinter in favor of blast furnace in Luleå, Sweden.26 Germany-based pelletized ore for charging blast furnaces: Because steelmaker thyssenkrupp has recently announced tests sinter produc on has a rela vely high emissions of hydrogen injec on in a blast furnace to subs tute intensity, reducing or elimina ng it in favor of for coal dust as a reductant.27 pelle zed ore signifi cantly reduces emissions.21 This ■ requires no capital investment or process changes for Improving hot blast stove operation: A number of blast furnace opera on and can improve overall blast techniques can be used to improve stove effi ciency, furnace performance.22 including air pre- or super-hea ng (see below), use of ceramic burners, and staggered airfl ow through ■ Substituting biomass-derived charcoal for coke mul ple stoves.7 in blast furnace charge, sintering and/or tuyere ■ injection: Charcoal derived from processes such Optimizing blast furnace operation using as autothermal pyrolysis of biomass can poten ally modeling and simulation with high-performance replace or substan ally subs tute for the use of coal- computing: Observing condi ons within the blast derived coke in blast furnaces, providing a source of furnace during opera on is diffi cult, so precise control carbon for combus on, structural strength to support over the process, including op mized injec on of charged materials and suffi cient porosity. Treated coal and charge, is not usually achieved. Simula ons biomass could also be used as a coke subs tute in using computa onal fl uid dynamics can improve the other steps of the BOF route. This would require understanding of blast furnace processes and lead to 28 investment in pyrolysis equipment and establishment process op miza ons. of a reliable biomass supply, but minimal or no ■ Plasma torch super-heating of hot blast air: changes to blast furnaces.23 Biomass subs tu on for Plasma torches use an electric arc to convert a coke has an es mated emissions reduc on poten al working gas to a plasma, achieving temperatures of of 32-58% for the BOF route.24 Based on a variety of approximately 5,000 °C. Introducing a plasma torch factors including the availability of sustainable sources using BFG and low-carbon electricity to superheat hot of biomass, the most suitable countries for biomass blast air could reduce the total coke use of the BOF subs tu on in iron and steelmaking are Canada, route.29 25 Sweden, China, the US and France. Process description: EAF and DRI ■ Modifi cation of the coking process: There are a variety of approaches to reducing emissions from In EAF steelmaking, a refractory-lined vessel is fi rst coking ovens, including single-chamber-system (SCS) charged with a mixture of steel scrap and DRI. Graphite coking—which uses single, large-volume ovens to electrodes within the vessel are powered with either AC achieve high thermal effi ciencies—and coke oven or DC electric current, and an electric arc forms between under-fi ring with advanced diagnos cs to improve the electrodes and the charge material. The material is hea ng effi ciency.9 heated both through resis ve hea ng from the passing electric current and radiant hea ng from the arc, ■ Reducing coke use through pulverized coal and which can reach temperatures of 3,000 °C. EAF vessels hot oxygen injection to the blast furnace: This range from a few tons to hundreds of tons per charge, technique subs tutes a frac on of blast furnace coke using transformers that range from 10 to 300 MVA consump on with direct coal injec on (bypassing and currents of up to 100 kA or higher. Depending on the coke-making process) and can achieve good furnace design and the charge proper es, approximately combus on with addi onal oxygen injec on.9 1.6 GJ of electrical energy is consumed per ton of melted ■ Reducing coke and coal use through co-injection steel.7,9 of hydrogen: The CO2 Ul mate Reduc on in

December 2019 38 EAFs are usually charged with scrap but can also accept avoiding the use of coke or the need to sinter ore. Ore DRI or even pig iron as part of the charge (see Figure and lower-grade steam coal are fed to the reactor, 3B-1). DRI is produced by reducing iron in solid form which is also able to accept a signifi cant frac on of (without mel ng), most commonly using a hydrogen- scrap and biomass. Tata Steel has demonstrated the carbon monoxide mixture (syngas) produced by process at a pilot plant in the Netherlands.32 reforming natural gas as a reducing agent; no coke is ■ Biomass-nugget smelting: US-based Carbontec required. Mul ple processes exist, but approximately Energy Corpora on is developing an alterna ve two-thirds of global DRI produc on (100.5 Mt in smel ng technology that packages ore, biomass and 2018) uses the MIDREX process, with HYL/Energiron limestone in compact “nuggets” (brique es) that are represen ng 15.5%.30 The emissions benefi ts come from then heated. This pyrolizes the biomass and leads two dis nct proper es of natural gas in ironmaking: to reduc on of the ore, without the use of coke or the emissions intensity per unit of thermal energy from sintering.33 combus on is nearly half that of coal, and methane as ■ Expanded use of hydrogen reduction of iron ore: a reducing agent is twice as eff ec ve as carbon on an Several processes, including HYBRIT and H2FUTURE, emissions basis.31 are being developed to use hydrogen as a reducing India manufactures the largest amount of DRI globally, agent for iron ore. The product would be DRI and, and primarily uses a process based on combus ng and when this is combined with EAF steelmaking using gasifying low-grade coal to produce CO as a reducing low-carbon electricity for hydrogen produc on from agent. The coal-based DRI process is suitable for smaller electrolysis, the resul ng process can have very low produc on units but has higher energy intensity. emissions intensity. When charging the EAF with 25% Mul ple types of reactors are used, including sha scrap, emissions from this route would be lower than furnaces, rotary kilns and fl uidized bed reactors.7 the BOF route for grid emissions intensi es below 661 34 kgCO2/MWh. Hydrogen reduc on has the advantage Medium- to long-term decarbonization of signifi cant opera onal fl exibility, since electrolyzers options involving significant process change can be ramped up and down to follow variable renewable genera on.35 A number of alterna ve iron-making processes ■ Electrolytic steel production: Many metals, have been developed and/or are in the process of including aluminum, zinc and nickel, are produced development. These involve substan al process changes at scale using electrolysis. Steel produc on has been and generally seek to reduce or eliminate the use demonstrated using this technique, with molten oxide of coke and blast furnace technology broadly. These as the electrolyte (which is able to withstand the approaches include the following: opera onal temperature of 1,600 °C) and carbon-free ■ Expanded use of smelting reduction for hot anodes. The emissions savings from this technique metal production: Smel ng reduc on processes fundamentally come from elimina ng a carbon- iron ore to hot metal in two stages. Ore is charged to based reductant (such as CO) and using electricity a reduc on sha reactor where process gases that directly instead. Technology developed at MIT and are produced from the melter-gasifi er reactor mostly commercialized by Boston Metal, as well as the reduce the ore to DRI. The DRI is then moved to the ULCOWIN process, present pathways to electroly c melter-gasifi er with discharge screws, and the fi nal steel produc on.9,36 reduc on and mel ng processes take place there. The two most commonly used processes are COREX and Integrating CCS into iron and steel making FINEX, the la er of which is able to charge ore without Carbon capture from fl ue gas at a BOF integrated agglomera on. No coke is required in the process, steel mill has substan al technical poten al for yielding important emissions savings.9 reducing overall emissions. Santos et al. examined ■ Upgraded smelting process: The HIsarna process three such scenarios: two based on post-combus on has been developed within the Ultra-Low Carbon capture technology using a conven onal solvent Dioxide Steelmaking (ULCOS) framework as an (monoethanolamine, MEA), applied to the fl ue gas from integrated hot metal produc on method. This process the hot stoves and steam genera on plant (resul ng

combines mul ple processes in a single cyclone oven, in 50.1% overall reduc on in CO2 emissions), with the

December 2019 39 Break-even cost Cost of CO Case Gas streams treated CO reduc on 2 2 ($/t HRC) avoided ($/t) Reference None 0% $575 N/A Post-combus on A Hot stoves, steam plant 50% $652 $74 Hot stoves, steam plant, coke Post-combus on B 60% $678 $81 oven, lime kiln OBF Blast furnace gas 47% $630 $57

Table 3B-1. CO2 emissions reduc on, break-even costs and cost of CO2 avoided for a BOF reference scenario and three carbon capture scenarios.37

higher-capture scenario also trea ng fl ue gas from the The ULCOS consor um has pursued the development coke ovens and lime kilns (resul ng in 60.3% overall of carbon capture from BFG through the concept of

reduc on in CO2 emissions); and a fi nal scenario based Top Gas Recycle-Blast Furnace (TGR-BF). This technique on an oxygen-blown blast furnace (OBF) in which replaces hot blast air with pure oxygen and recycles capture using a solvent blend (methyldiethanolamine, BFG a er capturing CO2 for re-injec on into the blast MDEA, with piperazine, PZ) is deployed on the BFG furnace; it has demonstrated savings of 20% or greater 39 (resul ng in 46.5% overall reduc on in CO2 emissions) in CO2 emissions. RIST and POSCO in South Korea have (see Table 3B-1).37 The use of carbon capture increases demonstrated the use of aqueous ammonia solvents and 40 the break-even price of steel produc on (without low-grade waste heat to capture CO2 from BFG. subsidies) from $575/t hot rolled coil (HRC) to $652/t The dominant DRI process, MIDREX, could poten ally HRC and $678/t HRC, respec vely, in the fi rst two cases accommodate CO capture from the slip stream of (post-combus on), and to $630/t HRC in the third case 2 recycled top gas using pressure swing adsorp on (OBF). These costs correspond to CO avoidance costs 2 (PSA). However, this would require addi onal process of $74/tCO , $81/tCO and $57/tCO , respec vely. 2 2 2 changes and has not been demonstrated. By contrast, The rela vely low value of the OBF case is primarily CO capture is an integral part of the exis ng HYL/ due to a signifi cant reduc on in coke consump on 2 Energiron process and is in commercial opera on at and corresponding emissions reduc on from the coke the Al Reyadah facility in the UAE, where captured ovens. In all three cases, cost increases were driven CO is used for off shore EOR. However, in general this by increased consump on of natural gas (due to 2 process only reduces overall emissions per ton of steel increased plant energy consump on, although energy consump on increase was almost negligible in the OBF scenario) and investment cost for capture equipment. Several research projects are underway to demonstrate CCS for BOF ironmaking. The COURSE50 program in Japan has focused on developing chemical and physical

adsorp on technologies for capturing CO2

from BFG, and is targe ng a cumula ve CO2 emissions reduc on of 30% when combined with co-injec on of hydrogen (see above, and Figure 3B-5).38 By developing new sorbent materials, the project has reduced the energy requirements for regenera on and has enabled the use of facility waste Figure 3B-5. The COURSE50 program in Japan has focused on developing heat (e.g., from slag, hot stove gas, etc.). chemical and physical adsorp on technologies for capturing CO2 from blast furnace gas. Credit: NEDO.

December 2019 40 by 25-35%, assuming average grid emissions intensity 4 Global steel industry: Outlook, challenges and for the associated EAF.37,41 The FINEX smel ng reduc on opportuni es, worldsteel (2017). process produces low-nitrogen tail gas that is compa ble 5 Chalabyan, A. The current capacity shake-up in steel with CO2 capture with no substan al process changes, and how the industry is adap ng, McKinsey & Company although the overall emissions reduc ons per ton of (2018). 37,42 steel are likely to be less than 50%. 6 Mabashi, D. Latest developments in steelmaking capacity, OECD (2019). Conclusions: Addressing process heat in iron 7 Carpenter, A. CO abatement in the iron and steel and steel making industry, IEA Clean Coal Center (2012). ■ Both the BOF and EAF routes require substan al 8 Li, Y. et al. CO emissions from BF-BOF and EAF amounts of process heat, but that heat is generated in steelmaking based on material fl ow analysis, Advanced very diff erent ways. Materials Research (2012). ■ EAF emissions are almost en rely due to electricity 9 Hasanbeigi, A. et al. Emerging energy-effi ciency and genera on (because the process is essen ally fully carbon-dioxide-emissions-reduc on technologies for electrifi ed), so the produc on of low-carbon process the iron and steel industry, LBNL (2013). heat is directly linked to the supply of low-carbon 10 Tolomeo, N. et al. US steel sector thrives as mills move electric power. up quality ladder, S&P Global Pla s (2019). ■ The BOF route generates heat by combus ng coke 11 Pauliuk, S. Steel all over the world: Es ma ng derived from metallurgical coal. The process is highly in-use stocks of iron for 200 countries, Resources, integrated, with the use of off -gases such as blast Conserva on and Recycling (2013). furnace gas (BFG) and coke oven gas (COG) for various 12 hea ng opera ons, including hot blast stove hea ng Gauffi n, A. et al. The global societal steel scrap reserves and amounts of losses. Resources (2016). and power genera on. It is therefore not prac cal to replace any individual hea ng unit opera on without a 13 Milford, R. et al. The roles of energy and material larger process redesign. effi ciency in mee ng steel industry CO targets, Environmental Science & Technology (2013). ■ There are a range of op ons to reduce emissions in the BOF route, including the use of biomass, plasma 14 Steel Sta s cal Yearbook 2018, World Steel. superhea ng of blast air, reducing sinter and hydrogen 15 Steel Sta s cal Yearbook 2010, World Steel. injec on. 16 Zhong, F. Is it me for China to switch to electric arc ■ Emerging iron-making processes such as smel ng furnace steelmaking? World Steel (2018). reduc on and direct-reduced iron (DRI) off er more 17 fl exibility in op ons to provide low-carbon heat. Demonstra on of a Waste Heat Recovery System at a Steel Plant in India, NEDO (2014). However, they are incompa ble with conven onal blast furnace technology and represent a substan al 18 Fernandez-Gonzalez, D. et al. Iron ore sintering: process change. Process. Min. Proc. Ext. Met. (2017). ■ CCS off ers a route to substan al emissions reduc on 19 AP42, Fi h Edi on, Chapter 12: Metallurgical Industry. from conven onal (BOF) steelmaking. Capture can be US Environmental Protec on Agency. applied to various fl ue gas streams at an integrated 20 Sarna, S. Genera on of hot air blast and hot blast steel mill or BFG. Some smel ng reduc on and DRI stoves. Ispat Guru website (2015). processes are highly amenable to carbon capture, 21 Ly on, W. Barriers to industrial decarboniza on. while others will require more research. Sandbag.org (2018).

22 1 The role of steel manufacturing in the global economy, Agrawal, A. Blast furnace performance under varying Oxford Economics (2019). pellet propor on. Trans. Ind. Inst. Met. (2019). 23 2 World Steel in Figures 2019, worldsteel (2019). Jahanshahi, S. et al. Recent progress in R&D on assessment of biomass/designer chars for steel 3 Mineral Commodity Summaries 2019, Iron and Steel, produc on, Innova on of Ironmaking Technologies and US Geological Survey (2019). Future Interna onal Collabora on Conference (2014).

December 2019 41 24 Mathieson, J. et al. Poten al for the use of biomass in 33 Fisher A. et al. U liza on simula on to guide furnace the iron and steel industry, Chemeca 2011 (2011). designs for the E-Iron nugget process. Lawrence Livermore Na onal Laboratory (2017). 25 Mandova, H. et al. Global assessment of biomass suitability for ironmaking – Opportuni es for 34 Vogl, V. Assessment of hydrogen direct reduc on for co-loca on of sustainable biomass, iron and steel fossil-free steelmaking, J. Cleaner Prod. (2018). produc on and suppor ve policies, Sustainable Energy 35 Ahman, M. et al. Hydrogen steelmaking for a low- (2018). carbon economy. Stockholm Environment Ins tute and 26 Nishioka, K. et al. Sustainable aspects of CO ul mate Lund University (2018). reduc on in the steelmaking process (COURSE50 36 Allanore, A. Features and challenges of molten oxide project), Part 1: Hydrogen reduc on in the blast electrolytes for metal extrac on, J. Elec. Soc. (2015). furnace. J. Sust. Met. (2016). 37 Santos, S. et al. Iron and steel CCS study (Techno- 27 Thyssenkrupp, “World fi rst in Duisburg as NRW economics integrated steel mill), IEAGHG, 2013/04, July economics minister Pinkwart launches tests at 2013. thyssenkrupp into blast furnace use of hydrogen”, Press Release, 11/11/2019; h ps://www.thyssenkrupp-steel. 38 Tonomura, S. et al. Concept and current state of com/en/newsroom/press-releases/world-fi rst-in- CO Ul mate Reduc on in the Steelmaking Process duisburg.html. (COURSE50) aimed at sustainability in the Japanese steel industry, J. Sust. Met. (2016). 28 Fisher, A. et al. Improving steel produc on with a virtual blast furnace, Lawrence Livermore Na onal Laboratory 39 Van der Stel, J. et al. ULCOS top gas recycling blast (2016). furnace process (ULCOS TGRBF): Final report, European Commission (2014). 29 Patel, N. et al. The use of plasma torches in blast furnace iron making, 46th Ironmaking Seminar (2016). 40 POSCO, Posco Report 2017. 30 2018 World Direct Reduc on Sta s cs, MIDREX (2019). 41 Sakaria, D. Case Study: Al Reyadah CCUS Project. Carbon Sequestra on Leadership Forum (2017). 31 Tanaka, H. Poten al for CO emissions reduc on in MIDREX direct reduc on process, Kobe Steel (2013). 42 POSCO, The FINEX Process: Economical and environmentally safe ironmaking (2015). 32 Fleschen, D. A possible future of steelmaking: HIsarna. Market Steel (2018).

December 2019 42 CHEMICALS Industry overview The global chemical sector (including petrochemicals and refi ning) is enormous and varied. Total sales in 2017 were $4 trillion.1 The chemical sector has grown rapidly over the past decade and is projected to con nue growth for many years.2 In 2017, the chemical sector

vented ~1.6 Gt CO2, roughly 3% of global CO2 emissions. Energy demand was greater for chemicals than for either cement or steel, refl ec ng enormous heat consump on (Figure 3C-1).3 The chemical sector represents the largest frac on of US Figure 3C-1. Global energy demand and direct (process) industrial produc on and GHG emissions. Chemical and CO emissions by sector, 2017. Source: IEA.3 refi ning industries respec vely emit 177 and 184 million 2

tons CO2e annually—roughly 25% of total US industrial produc on processes is extremely varied, including CO2 emissions—with 28% concentrated in Texas and Louisianaa. To achieve key climate goals, prac cal the Haber-Bosch process (ammonia), methana on approaches to decarbonizing chemical produc on is of (methanol produc on), ethylene cracking (ethylene) vital importance. and pyrolysis of heavy crude (carbon black produc on). Many of these reac ons require fi t-for-purpose reactors Unlike cement or steel produc on, which produce a that cannot be readily replaced. small number of commercial products, the chemical sector produces an enormous variety of diff erent In general, many chemical applica ons require heat that products using diff erent processes. These include is rela vely low in temperature (300-600 °C) compared drop-in fuels (e.g., gasoline, diesel, jet fuel, bunker to other industries. This could open more poten al fuel), rela vely simple feedstocks and compounds sources of low-carbon heat to applica on, including (e.g., ethylene, methanol) and complex products (e.g., advanced nuclear systems or advanced electrical hea ng lubricants, carbon fi ber). The variety of produc on (e.g., induc on or radia ve hea ng approaches). This methods and products complicates strategies for is par cularly true for systems that require steam as a decarboniza on. primary feedstock (e.g., ethylene cracking), since boilers and steam generators today operate on a wide range Similarly, most chemical facili es have a wide set of of heat fuels and inputs. While the temperature range feedstocks and fuels. Refi neries consume natural for chemical applica ons is lower than for cement, steel gas, natural gas liquids and heavy hydrocarbons (e.g., and glass, the temperatures are generally too high for bitumen, asphalt) which serve as feedstocks as well as conven onal low-carbon heat systems (i.e. heat pumps). fuels. Hydrogen is also an important feedstock. In the US and EU, hydrogen produc on and dedicated pipelines A concern specifi c to the chemical industry is the broad e directly to refi ning and chemical facili es. In OECD distribu on of heat sources. Unlike in cement and steel, countries, natural gas is the predominant heat fuel and where most of the emissions fl ow from one or two large feedstock, while in China, India and Southeast Asia, coal reactors (e.g., a kiln or blast furnace), it is unusual to provides a large frac on of both heat and hydrogen. fi nd a single large source in most chemical plants—there are excep ons, e.g. cataly c crackers for fuel refi ning Complexities of Operation and synthesis. Large chemical plants have dozens or even hundreds of small emissions sources ed to heat The complexity of the chemical sector is matched by the produc on, including burners, furnaces and boilers. complexi es of facility opera on. The range of chemical Decarbonizing these small, distributed sources would likely add complexity and cost to plans for heat supply subs tu on. As such, poten al subs tutes for low- a See h ps://www.epa.gov/ghgrepor ng/ghg-repor ng-program-data-sets

December 2019 43 carbon heat may be limited by the breadth and diversity US, 12.5 million tons of ammonia worth $13.2 billion of heat consump on in a plant. Moreover, some large came from 34 facili es in 2018, mostly concentrated process units (e.g., cataly c crackers, carbon black in Texas, Oklahoma and Louisiana (in large part due to synthesis units) generate heat by par al oxida on and low-cost natural gas and associated low-cost hydrogen combus on of feedstocks and fuels in the key reac ons. produc on), resul ng in 11.8 million tons CO2 annual This limits the kind of subs tu ons that are possible, emissions.6,8 Annual market growth in the US is about since the core process reac ons require carbonaceous 2.5%, and global annual growth closer to 4%. Some fuels and feedstocks. recent analysis has proposed ammonia as either a hydrogen storage and transporta on medium or as a To limit our inves ga ons and focus on representa ve poten ally carbon-free fuel op on.9 If either op on sectors, we selected two pathways for discussion: entered the market at scale, there would be drama c ammonia, which chemically contains no carbon, and increase in ammonia produc on. methanol, which does. Both are large sub-sectors of the chemical industry. Both are globally-traded commodi es Almost all ammonia produc on worldwide uses the with mature supply chains and technologies. In this way, Haber-Bosch process, invented and developed largely both pathways are representa ve of the chemical sector from 1900 to 1910.10 The core process involves breaking as a whole, and specifi c issues and processes within the triple bond of nitrogen gas (N2) and reducing it by each provide some insight into both the challenges and adding hydrogen to form ammonia (NH3). The triple poten al solu ons to decarbonizing heat and produc on bond is very strong, so much energy is required to across other chemical supply chains. undertake ammonia synthesis. The process operates at very high pressure (2200-3600 psi/15-25 MPa), Ammonia which requires a lot of compression and mechanical Ammonia is a huge global industry. It underlies almost work, expensive capital equipment, and a fairly high all fer lizer produc on (450 million tons, $156 billion temperature (400-500 °C), which requires substan al b market) and is the largest single commodity within that heat . market, roughly 170 Mt worth $50 billion in 2017.6 b Global CO2 emissions from ammonia are roughly To learn more about the Haber-Bosch Process, see: h ps://www. 1.5% of global emissions (~490 MT in 2012).7 In the sciencedirect.com/topics/engineering/haber-bosch-process (Science Direct, 2019, The Haber-Bosch Process) BOX 33 Synthe c fuels and chemicals produc on in China and India A very large share of today’s chemical market is in China, India and Southeast Asia.2 In these regions, access to natural gas as a feedstock is o en limited, especially in China, India, Vietnam and Indonesia. In contrast, coal and petcoke are o en abundant and very cheap. This has led many chemical facili es in these countries to use solid carbon fuels for heat and also for hydrogen produc on. Many of these individual facili es are very large and clustered in key regions (e.g., Ningxia and Ahmedabad)4,5 and would require very largerge supplies of low-carbon heat and feedstocks to decarbonize. One feature of these facili es is that hydrogen produc on commonly occurs throughgh coal or petcoke gasifi ca on combined with water-gas shi . This produces a very

large byproduct (process) stream of concentrated CO2, which could be captured and stored for rela vely low cost. Unfortunately, none of the countries in ques on

have announced or executed serious plans to capture and store CO2. Also, in many of these regions, water is scarce and agriculture is limited. As such, solid biomass or biomethane subs tu on for heat is not a likely op on today.

December 2019 44 Figure 3C-3. Simplifi ed cartoon of Haber-Bosch process.

The power and heat requirements have driven effi ciency mul ple steps for ammonia synthesis require addi onal improvements industry wide, and the industry has heat to provide energy to upgrade/reduce nitrogen reduced energy requirements 75% since 1930 and during each run across cataly c beds (Figure 3C-4). roughly 40% since 1970.11 Global average footprint of ammonia produc on is es mated to be 1.9 tons CO2/ton Methanol ammonia, with ammonia made from natural gas having Like ammonia, methanol is a huge global industry and 12 a footprint of 1.1 tons CO2/ton ammonia. Over 50% is interna onally traded commodity. It is used both as a associated with hydrogen produc on and roughly 30- fuel and as a feedstock, usually for plas cs, fuels or more 40% from process heat. Hydrogen is a cri cal feedstock complex chemicals (e.g., formaldehyde, gasoline and to ammonia synthesis, and roughly 50% of global dimethyl ether). The global methanol market produced hydrogen produc on goes into ammonia produc on. 110 million tons worth $24.7 billion14 at over 90 plants Hydrogen is commonly made from natural gas or c worldwide . Global CO2 emissions from methanol are longer hydrocarbons at high temperature, which itself ~100 Mt in 2016.15 About 12 million tons16 of methanol consumes much heat during produc on. (See Chapter worth ~$3.6 billion17 came from US facili es, mostly in 2A). The hydrogen and nitrogen gases run over mul ple Texas and Louisiana. Global annual growth is closer to beds with iron-based catalysts followed by cooling 7.7% and is projected to grow at an 11% rate, reaching (and hea ng) in between. Each catalyst executes one $48 billion in 2024. This demand will be met in part from component of the chemical reac on, and the mul ple new produc on and export facili es in the US.17 beds are used to increase selec vity and yield. Originally, methanol was derived from wood and was The major consumers of heat are hydrogen synthesis known as wood alcohol. Today, methanol is typically (steam-methane reforming [SMRs]) and the ammonia made by feeding syngas into a methanol synthesis synthesis reactor. Because heat is a major cost to unit followed by dis lla on (Figure 3C-5). Both CO ammonia produc on due to mul ple hea ng and and CO2 can be converted to methanol, although CO2 cooling steps, most facili es a empt to capture waste heat through a set of heat exchangers (Figure 3C-3). The c See h ps://www.methanol.org/the-methanol-industry/

December 2019 45 Figure 3C-4. Schema c Gibbs free-energy chart showing all process steps in ammonia synthesis. Although the reac on is exothermic and theore cally yields energy, thermodynamic losses require addi onal energy input. Source:13

conversion requires more hydrogen than CO. Commonly, be replaced with hydrogen burners, electrical hea ng the reac on produces addi onal hydrogen, which can or steam produced by any carbon-free heat, including be sold or used inside the facility for heat or power. conven onal nuclear reactors. Because of the wide Methanol synthesis is insensi ve to the source of range of poten al pathways to generate heat, methanol the primary chemicals, which could be derived from facili es could serve as a poten ally important tes ng

captured CO2, biofuels, green hydrogen or others. ground for alterna ve heat supply subs tu on. There are many alterna ve produc on methods for methanol synthesis,18 including biomass-derived Decarbonization pathways for chemicals and syngas and heat,19 direct electroly c produc on, related heat 20 Several poten al pathways exist that could meaningfully waste conversion and synthesis from recycled CO2, 21 and green hydrogen. The George Olah plant in contribute to deep reduc on in chemical CO2 Icelandd uses this last methodology. In some cases, no emissions. These include demand destruc on, material process heat is required; although, for the electrical subs tu on, electrifi ca on, carbon capture and storage pathways, substan al amounts of addi onal energy are (CCS) and fuel switching. Es mates for poten al size required.22,23 Also, unlike ammonia, which burns without and pace of contribu on to decarboniza on are returning carbon to the atmosphere, methanol use highly controversial and some mes involve full facility ul mately returns the chemical-embodied carbon to

the air and oceans. If recycled CO2 is not provided by low-carbon biomass or CO2

captured from the air, it will add CO2 to the air and oceans. In contrast to many of the other sectors and approaches discussed in this report, methanol synthesis and dis lla on operate at fairly low temperatures (~300 °C). Small burners and furnaces commonly provide heat, which hypothe cally could

d See h ps://www.carbonrecycling.is/george-olah Figure 3C-5. Schema c cartoon of a methanol synthesis plant.

December 2019 46 replacement or changes in commercial prac ces and consumer preferences. As before, this discussion focuses on low-carbon fuel subs tu on for heat and retrofi ts to exis ng facili es.

Short-term decarbonization options involve gaseous substitutes The most common heat source for chemicals is natural gas. Most facili es are situated near major natural gas hubs (current or historical), and supply chains provide natural gas both as a feedstock and for heat. This means that the simplest pathways to subs tute low-carbon heat Figure 3C-6. Scenario-based direct CO2 emissions reduc ons in both process- and energy-related emissions for the chemical sector. Source: IEA.11 systems are with low-carbon gas supplies.

For these reasons, biomethane is the commercial applica on of electrical or biofuel-provided simplest op on, as it can immediately subs tute into heat, other outcomes are possible given diff erent exis ng plants with near-zero modifi ca on of the system assump ons. (e.g., 24,25). For this to yield low-carbon heat, the life- cycle footprint of the biomethane must be low, whether NOTE: Because many chemical products have embedded supplied by gasifi ca on, landfi ll or digester. It is not clear carbon (e.g., methanol) which re-enters the air and if this is likely to prove compe ve in chemicals—the oceans a er use, truly deep decarboniza on may IEA Clean Transi on scenario for chemicals found limited ul mately require radically diff erent prac ce than is uptake of biofuels and biomass substa on.12 used in today’s industry. Either all embedded carbon must come from recycled CO2 (e.g., DAC-supplied or Hydrogen, blue or green, is also a viable op on and is low-carbon-biomass-supplied CO2; biopolymers) or the most straigh orward to assess in terms of life-cycle alterna ve processes that do not emit their carbon analysis (LCA). The temperature of hydrogen burned in content a er use will be required.26 air is more than suffi cient for all chemical and refi ning processes. Most systems can accommodate 7-20% A number of alterna ve processes for ammonia, hydrogen blends based on the specifi cs of facility methanol and other chemicals have been developed engineering, and indeed some chemical plants already or are under development. These involve substan al use byproduct hydrogen (e.g., from ethylene synthesis) process changes and generally seek to reduce or to produce heat. All of the challenges discussed in eliminate the use of fossil-fuel feedstocks. the hydrogen sec on (e.g., embri lement, corrosion, One approach is electrosynthesis of feedstocks, including specifi c sensors and controls) would apply to such a hydrogen and syngas, which can be burned for heat and system. fuel. Both ammonia and methanol can be converted to electricity through a fuel cell to generate electricity and Medium- to long-term decarbonization options byproduct N or CO2. Using a reverse fuel cell, methanol involving significant process change 2 can be generated by adding CO2 and electricity to a Unlike blue hydrogen or biomethane, other pathways to fuel cell in an aqueous environment or the presence decarbonizing refi ning and chemical produc on are not of hydrogen. Haldor-Topsoe27 has designed and piloted yet commercial and in many cases are not yet piloted. this approach, which generates a syngas of hydrogen However, combina ons of effi ciency, heat subs tu on and nitrogen which enters a conven onal Haber-Bosch and CCS can provide substan al reduc ons in energy- reactor. This approach is similar to the JGP process related emissions—roughly 50% (Figure 3C-6). While piloted at Fukishima28 and does not require heat for this set of scenarios fi nds li le decarboniza on through SMR produc on or energy for the air separa on unit.

December 2019 47 The process would s ll require heat for the ammonia Integrating CCS into chemical processes synthesis reac ons, which could be supplied through For many chemical manufacturing processes, CCS can hydrogen or ammonia combus on. signifi cantly and cost-eff ec vely reduce CO2 emissions. According to IEA analysis,3 CCS is the chief op on Electrification of heat of many expected to contribute to least-cost GHG Prior sec ons discussed the poten al for electrical reduc on in the chemical sector, followed by effi ciency methods (e.g., resistance or dielectric hea ng) to and the switching of coal to gas. This result is both provide thermal energy to industrial processes. In the robust and unsurprising. Many produc on pathways case of chemicals, some electrifi ca on will likely prove (ethanol, methanol, ammonia) have large byproduct CO2 straigh orward and simple. For example, process streams. Successful introduc on and deployment of CCS steam used for chemical synthesis could be supplied by in chemicals will require investment and engineering electrical water heaters with limited capital expenditure. that integrate CO2 capture into conven onal opera ons The same is true for some small electrical furnaces. of chemical facili es. As elsewhere in this report, we Such systems exist today at 0.5-4 MW ra ngs and could focus on applica on to exis ng facili es, as opposed to supply distributed steam and heat. However, there new processes or integrated designs for new facili es. have been no published examples of facili es that have Carbon capture, use and storage (CCUS) applied to ins tuted such electrifi ca on nor has there been an pure streams of CO already operate today. Four large industry census regarding the extent to which these 2 hydrogen plants, one small ammonia plant, one refi nery replacements could occur in a straigh orward manner. and one ethanol plant operate today with CCS.29 All For more complex chemical reactors (e.g., ammonia or of these plants capture process emissions from pure ethylene synthesis), commercial electrical reactors are byproduct streams. However, CCS can also contribute not commercially available. Given the dimensions and substan ally to reducing heat-related emissions. In opera onal requirements of commercial facili es, it does ammonia produc on, for example, the primary process not appear possible to retrofi t exis ng reactors with emissions come from SMR hydrogen produc on. If a electrical hea ng methods today. In addi on, extremely dedicated hydrogen SMR plant was suffi ciently large low fi rm power costs are required to displace other to produce hydrogen for both ammonia synthesis and low-carbon heat op ons like biomass or CCS (see below: reactor heat, it could achieve 85-90% CO2 emissions costs), and displacement costs may vary as a func on of reduc on by turning the plant “blue”—adding CCS to the facility age, design and the replacement costs compared SMR facility and using the addi onal hydrogen for heat. to retrofi t costs. In essence, this could decarbonize both process and

Figure 3C-8. Cost es mates for diff erent heat decarboniza on pathways for ammonia and methanol.

December 2019 48 heat emissions through pre-combus on

separa on of CO2 from fuels.

Post-combus on CO2 separa on may be another method capable of reducing emissions at refi ning and chemical plants. This may prove most a rac ve for facili es that have large point sources (e.g., cataly c crackers, steam boilers, central furnaces) within their fence lines. In general, post-combus on capture will be more complicated, chiefl y due to the range and distribu on

of CO2 sources associated with both process emissions and heat. While it may be theore cally possible to retrofi t dozens of small sources for capture, it is likely to prove challenging and possibly infeasible. In either pre- or post-combus on applica ons, CCS systems will Figure 3C-9. Es mated abatement costs for greenfi eld ammonia produc on require transporta on and storage as a func on of electricity cost compared to biomethane and blue hydrogen infrastructure. Many poten al (CCS on natural-gas SMR). Source: McKinsey (2018)31 facili es for CCS retrofi t (pre- or post-combus on) lack pipeline capacity to transport industrial sectors, including for ammonia and methanol CO2 and lack ready and available storage sites. Those opera ng today are largely bespoke contracts for EOR. as proxies for the chemical sector (Figure 3C-8). For Although projects have been proposed for CCS clusters ammonia, they es mate a 5-40% increase in ammonia and hubs in Europe (e.g., PORTHOS and Teeside),30 produc on costs for most low-carbon heat pathways, the infrastructure has not yet materialized for lack with green hydrogen subs tu on delivering substan ally of investment. For CCS to contribute substan ally to higher unit produc on costs (60-120% increase). For decarboniza on of the chemicals industry, investment in methanol, they es mate 5-80% increase (with green key infrastructure will prove essen al. hydrogen increasing costs 125-190%). In the case of methanol, CCS to heat or to the whole facility appeared Cost estimates to be the lowest cost op ons. Es ma ng costs of decarbonizing heat supplies in the These es mates are in line with other published 31 chemicals sector remains diffi cult. In part, this is due to es mates. For example, McKinsey (2018) es mated the range of op ons and the diffi cul es in es ma ng that carbon-free ammonia would increase unit price the true carbon footprints of viable op ons. It is also ~5 to 35 percent depending on the future price of partly is due to factors that are simply diffi cult to assess renewable electricity and that power costs would or forecast in the present or future. The es mates need to range from $25-45/MWh to be compe ve discussed here have large ranges and substan al with other low-carbon synthesis pathways (Figure uncertain es. In all cases, however, alterna ve 3C-9). They considered this economic hurdle to be approaches to decarbonizing heat add substan al cost to substan al enough that “decarboniza on would require unit produc on. technological breakthroughs, a further lowering of zero- carbon energy prices, changing customer preferences Using specifi c assump ons for power price and (willingness to pay) and/or a regulatory push.” This is availability, natural gas price, and other factors, similar to the conclusion of Abanades et al.22 or IEA’s Friedmann et al. (2019) produced es mates for many

December 2019 49 (2013) es mates32 of addi onal energy requirements for renewable chemical synthesis (Figure 3C-10).e Finally, the Mission Possible report33 explained the addi onal required energy to supply only methanol or ammonia as a shipping fuel through electrical synthesis: “Total electricity genera on, whether for direct use, or for the produc on of hydrogen, ammonia or synthe c fuels, will need to grow from around 20,000 TWh today to 85-115,000 TWh by mid-century. This hugely increased electricity supply will have to be produced at 85-90% from direct zero-carbon electricity genera on (i.e. renewables or nuclear) with only 10- 15% coming from biomass or abated Figure 3C-10. Addi onal energy requirements for subs tu on of renewable fossil fuel inputs.” power for hydrogen-based chemical synthesis. Percentages represent frac ons of total global market. Source: IEA (2013)32 This framework requires system costs for fi rm zero-carbon power to be between lower costs or more advanced technology to be $5-25/MWhr—an extremely low power price—for such compe ve and scalable. fuels to be cost compe ve absent policy support. ■ Alterna vely, it will be essen al to fi nd new and Carbon capture and storage (CCS) will likely prove innova ve pathways that require much less addi onal important to decarbonizing chemical produc on, energy or electricity to produce commodity chemicals. either in the manufacture and use of “blue hydrogen” or to decarbonize fl ue gas post-combus on. Conclusions: Addressing process heat in 1 M. Garside, “Total revenue of the global chemical chemical manufacturing industry from 2002-2017,” Sta sta (2019), h ps:// ■ Chemical produc on, including refi ning, emits www.sta sta.com/sta s cs/302081/revenue-of-global- chemical-industry/ substan al CO2 from heat—roughly 50% of the emissions from chemical produc on result from 2 Oxford Economics, 2019, The Global Chemical Industry: produc on of heat. Catalyzing Growth and Addressing Sustainability ■ Almost all heat used in chemical produc on is Challenges, 28 p. h ps://www.icca-chem.org/ provided by natural gas. This suggests that the easiest wp-content/uploads/2019/03/ICCA_EconomicAnalysis_ Report_030819.pdf (near-term) subs tu on op ons come from low- carbon gas fuels, including decarbonized hydrogen and 3 Interna onal Energy Agency (IEA)/Organiza on for biogas. Economic Co-opera on and Development (OECD), “The Future of Petrochemicals: Towards a more sustainable ■ Medium- and long-term approaches require much plas cs and fer lisers—Execu ve Summary,” IEA Publica ons (2018), h ps://webstore.iea.org/ download/summary/2310?fi leName=English-Future- e From Katelhon et al.’s analysis of full subs tu on of electrochemical Petrochemicals-ES.pdf routes: “If all addi onal electricity were provided by renewable energy, the amount of renewable energy required for the full-scale 4 introduc on of CCU would correspond to 126% and 222% of current Xing Zhang, “Visi ng the World’s Biggest Single targets [sustainable development scenario of IEA (31)] for the global Coal-to-Liquid Project in Yinchuan, China,” IEA Clean renewable electricity produc on in 2030 for the low-TRL and the high-TRL Coal Centre (2017), h ps://www.iea-coal.org/ scenario, respec vely… the need for a further expansion of renewable electricity produc on capaci es is likely to be a limi ng factor for CCU visi ng-the-worlds-biggest-single-coal-to-liquid-project- in the chemical industry.” (h ps://www.pnas.org/cgi/doi/10.1073/ in-yinchuan-china/ pnas.1821029116)23

December 2019 50 5 “India’s RIL coke gasifi ers aid petrochemical strategy,” 15 Zachary J. Schiff er and Karthish Manthiram, Argus Media Group (2019) h ps://www.argusmedia. “Electrifi ca on and Decarboniza on of the Chemical com/en/news/1958372-indias-ril-coke-gasifi ers-aid- Industry,” Joule, (2017), h ps://www.sciencedirect. petrochemical-strategy com/science/ar cle/pii/S2542435117300156. 6 USGS, 2019, 2016 Minerals Yearbook: Nitrogen, 16 ADI Analy cs, LLC, “Economic and Employment h ps://prd-wret.s3-us-west-2.amazonaws.com/assets/ Impacts of U.S. Methanol Industry,” Methanol Ins tute palladium/produc on/atoms/fi les/myb1-2016-nitro.pdf (2017), h p://www.methanol.org/wp-content/ uploads/2017/06/Uday-Turaga-Domes c-Methanol- 7 Ammonia Industry, 2016, Ammonia produc on causes Industry-Resurgence-1.pdf 1% of total global emissions, h ps://ammoniaindustry. com/ammonia-produc on-causes-1-percent-of-total- 17 Marc Alvarado, “Methanol,” Methanol Ins tute global-ghg-emissions/ (2016), h p://www.methanol.org/wp-content/ uploads/2016/07/Marc-Alvarado-Global-Methanol- 8 US EPA, 2009, Technical Support Document for February-2016-IMPCA-for-upload-to-website.pdf the Ammonia Produc on Sector: Proposed Rule for Mandatory Repor ng of Greenhouse Gases, 18 “Methanol Produc on,” Methanol Ins tute Offi ce of Air and Radia on, h ps://www.epa.gov/ (2016), h p://www.methanol.org/wp-content/ sites/produc on/fi les/2015-02/documents/ _g- uploads/2016/06/MI-Combined-Slide-Deck-MDC- tsd_ammonia_epa_1-22-09.pdf slides-Revised.pdf 9 Cédric Philibert, “Renewable Energy for Industry: 19 BioMCN, “Sales Specifi ca on Bio-Methanol,” (2014) From green energy to green materials and fuels,” In 20 Arlene Karidis, “Enerkem to Make Methanol Through Insight Series 2017 – Renewable Energy for Industry, Gasifi ca on in Netherlands,” Waste360 (2018), h ps:// (2017), h ps://webstore.iea.org/insights-series-2017- www.waste360.com/waste-energy/enerkem-make- renewable-energy-for-industry methanol-through-gasifi ca on-netherlands 10 Smil, Vaclav. “Enriching the Earth: Fritz Haber, 21 Khalid Mubarak and Rashid Al-Hitmi, “QAFAC: Carbon Carl Bosch, and the Transforma on of World Food dioxide recovery plant,” QScience (2012), h ps:// Produc on.” MIT Press (2001). p. 358. ISBN 978-0-262- www.qscience.com/docserver/fulltext/stsp/2012/2/ 69313-4. stsp.2012.ccs.22.pdf 11 The European Chemical Industry Council (CEFIC), 22 J. Carlos Abanades, et al., “On the climate change “European chemistry for growth: Unlocking a mi ga on poten al of CO conversion to fuels,” Energy compe ve, low carbon and energy effi cient future,” & Environmental Science (2017), v.10, 2491, DOI: CEFIC (2013), h ps://cefi c.org/app/uploads/2019/01/ 10.1039/C7EE02819A Energy-Roadmap-The-Report-European-chemistry-for- growth_BROCHURE-Energy.pdf 23 Kätelhön, et al., “Climate Change Mi ga on Poten al of Carbon Capture and U liza on in the Chemical 12 Interna onal Energy Agency (IEA), “The Future of Industry.” Proceedings of the Na onal Academy Petrochemicals: Towards more sustainable plas cs and of Sciences (2019). h ps://doi.org/10.1073/ fer lizers (Full Report)” IEA (2018), h ps://www.iea. pnas.1821029116. org/petrochemicals/ 24 Energy Futures Inita ve, 2019, Op onality, Flexibility 13 G. Ertl , Surface Science and Catalysis: ”Studies on the & Innova on: Pathway for Deep Decarboniza on in Mechanism of Ammonia Synthesis: The P. H. Emme California (May 2019), h ps://sta c1.squarespace. Award Address. (1980) In: Catalysis Reviews. 21, com/sta c/58ec123cb3db2bd94e057628/t/5ced S. 201–223, doi:10.1080/03602458008067533 6fc515fcc0b190b60cd2/1559064542876/EFI_CA_ 14 “Methanol Market: Global Industry Trends, Share, Size, Decarboniza on_Full.pdf Growth, Opportunity and Forecast 2019-2024,” h ps:// 25 Materials Economics, 2019, Industrial Transforma on www.researchandmarkets.com/reports/4763079/ 2050: Pathways to net-zero emissions from EU heavy methanol-market-global-industry-trends-share?utm_ industry, 208p. h ps://materialeconomics.com/ source=GN&utm_medium=PressRelease&utm_ publica ons/industrial-transforma on-2050 code=xslpqj&utm_campaign=1237553+-+The+Global+ Methanol+Market+is+Projected+to+Exceed+%2448+Bil 26 Sandalow, David, Roger Aines, Julio S. Friedman, Colin lion+by+2024+--+China+Enjoys+the+Leading+Posi on& McCormick, and Sean T. McCoy, “Carbon Dioxide utm_exec=joca220prd U liza on (COU): ICEF Roadmap 2.0.” Tokyo, Japan: Innova on for Cool Earth Forum (2017), h p://www.

December 2019 51 icef-forum.org/pla orm/upload/COU_Roadmap_ 30 European Commission (EC), “The poten al for CCS and ICEF2017.pdf. CCU in Europe. Report to the thirty second mee ng of the European Gas Regulatory Forum 5-6 June 27 John B. Hansen, Pat A. Han (Haldor Topsøe), “Roadmap 2019”, European Gas Regulatory Forum, Madrid, Spain to All Electric Ammonia Plants,” NH3 Fuel Associa on, Organized by the Interna onal Associa on of Oil and 15th Annual NH3 Fuel Conference, Pi sburgh, PA Gas Producers (IOGP), (2019) h ps://ec.europa.eu/ (2018), h ps://nh3fuelassocia on.org/2018/12/07/ info/sites/info/fi les/iogp_-_report_-_ccs_ccu.pdf roadmap-to-all-electric-ammonia-plants/ 31 Mckinsey, 2018, Decarboniza on of the industrial 28 Hideyuki Matsumoto, et al., “Analysis of infl uence sector: the next fron er. h ps://www.mckinsey.com/ of opera ng pressure on dynamic behavior of industries/oil-and-gas/our-insights/decarboniza on-of- ammonia produc on over ruthenium catalyst under industrial-sectors-the-next-fron er high pressure condi on,” NH3 Fuel Associa on, 15th Annual NH3 Fuel Conference, Pi sburgh, PA 32 Interna onal Energy Agency (IEA), “Technology (2018), h ps://nh3fuelassocia on.org/2018/12/14/ Roadmap: Energy and GHG reduc ons in the Chemical analysis-of-infl uence-of-opera ng-pressure-on- Industry via Cataly c process,” IEA (2013), h ps://www. dynamic-behavior-of-ammonia-produc on-over- iea.org/publica ons/freepublica ons/publica on/Tech ruthenium-catalyst-under-high-pressure-condi on/ nologyRoadmapEnergyandGHGReduc onsintheChemic al-IndustryviaCataly cProcesses.pdf 29 Global CCS Ins tute (GCCSI), “Global Status Report,” GCCSI (2019), h ps://www.globalccsins tute.com/ 33 Energy Transi on Commission, 2018, “Mission Possible: resources/global-status-report/ Reaching net-zero carbon emissions from harder-to- abate sectors by mid-century, Report,” h p://www. energy-transi ons.org/mission-possible

December 2019 52 there are 100,000 miles of pipeline feeding residen al, CHAPTER 4 commercial and industrial clients. Currently a small amount of renewable natural gas is being put into the gas systems in Europe and the US. INNOVATION Created mainly from anaerobic diges on of waste, like sewage and agricultural waste, using this gas creates a PATHWAYS nearly carbon-free combus on opportunity. Denmark’s Innova on in industrial heat is one of the more diffi cult gas grid has about 50% biomethane content and is topics in climate change mi ga on, in part due to the projected to be 100% renewable by 2035. This is a very large number of processes that must be improved. high percentage, refl ec ng an availability of manure Industrial heat is deeply embedded in our economy and bioenergy crops that is unlikely to be met in other and implemented in diverse processes. It is much less countries. Recent es mates of the renewable natural gas suscep ble than electricity or transporta on to major capacity of California, for instance, place the total at no changes in the way energy is delivered, rather than how more than 20% of today’s total while using only waste it is used. The use of hybrid approaches, such as carbon resources (there is no current energy crop contribu on capture and storage (CCS) with par al biomass use to to renewable natural gas in California). improve the overall carbon footprint, may be vital to As electrifi ca on proceeds in homes and commercial successful decarboniza on of industrial heat. facili es, the total amount of natural gas used will drop, In addi on to the specifi c op ons outlined in the making it possible for a higher percentage of the total previous chapters, four very diff erent innova on to be renewable. By adding power-to-gas systems that pathways appear likely to be broadly useful for a variety convert electricity into either hydrogen or methane, of industries: na ons may be able to provide a large propor on of the industrial need for gas via renewable sources. Current 1. Revising the fuel mix to provide low-carbon- or zero- es mates of the allowable percentage of hydrogen in carbon-footprint fuels without major changes in the a natural gas system vary widely, depending on issues industrial process. of corrosion. In 2018, the GRHYD project in France 2. Improving the way heat is applied in processes, (Ges on des Réseaux par l’injec on d’Hydrogène pour including heat storage. Décarboner les énergies [grid management through 3. Hybrid approaches, including CCS with process the injec on of hydrogen for energy decarboniza on]) improvements and nega ve emissions technologies began blending 6% hydrogen into the natural gas grid that remove the CO2 at sites remote from the and will test up to 20%.1 Also, whether hydrogen can be industrial facility. stored in geologic natural gas sites is not known. An R&D 4. Cross-cu ng systema c changes in hydrogen and eff ort to determine these limits is needed. A signifi cant biomass supply that address mul ple industry limita on on this op on is the maintenance cost of the sectors. gas grid, which presumably must be shared among fewer users while not changing the more expensive aspects of Delivering zero-carbon fuels via existing the system. A transi on plan for such a future must be infrastructure developed. Gas If hydrogen at high levels is not permissible in exis ng gas systems, it could be converted into methane by Today’s natural gas system powers much of the most reac on with CO2. The Saba er process is in use in a effi cient industries in the developed world. Most demonstra on facility that Audi operates in Germany discussions of a completely decarbonized world to make renewable natural gas.2 An issue with this assume that we stop using natural gas and accordingly approach is that it creates heat from the exothermic abandon that infrastructure. However, in many areas reac on that must be used if the overall energy balance the gas distribu on grid is a massive distribu on system of the system is to be reduced. Several research groups already in place—for instance, in the Los Angeles area are also pursuing biological systems to directly convert

December 2019 53 electricity into methane via microbial popula ons, led H2 produc on (e.g., Sulfur-Iodine cycle discussed by Germany’s Electrochea.3 They are conduc ng a small- previously). Hydrogen produc on may become global scale demonstra on at the Na onal Renewable Energy as well, with both Australia6 and the UAE considering Laboratory in the US, with promise for direct conversion large-scale export of hydrogen by refrigerated tanker. of renewable energy into methane. This technology uses France announced its Hydrogen Deployment Plan for hydrogen and CO2 as the feed for microbes that convert Energy Transi on in June 2018, the targets of which the gases into methane. When using electrochemical include 20-40% low-carbon hydrogen use in industrial genera on of hydrogen, the process is about 50% applica ons of hydrogen and a reduc on in electrolysis energy effi cient at conver ng electricity into methane. cost €2-3/kg by 2028.1 This innova on pathway will Obviously, this s ll requires an effi cient carbon-free require considera on of the needs for infrastructure hydrogen source. to accommodate that scale of import, but since many major industrial facili es are near ports, this could be a Like the Saba er process, near-term methods of u lizing drama c change in world energy markets. the exis ng gas grid involve hydrogen, either directly or as fuel to make other energy carriers, such as methane An innova on pathway of major hydrogen produc on, and ammonia. New means of crea ng hydrogen are also from zero-carbon natural gas conversion or from in development beyond the green and blue hydrogen renewable energy followed by liquifi ed hydrogen men oned earlier. Methane pyrolysis, where hydrogen transport over the high seas, could drama cally change is stripped from methane at temperatures of 800-1100 the future of industrial heat sources but will require °C, is a promising technique that leaves solid carbon as a massive new set of infrastructure for hydrogen a byproduct or waste to be landfi lled.4 Two methods genera on and transport, while allowing exis ng are currently in considera on. In the older method, industry to keep very similar processes and procedures. methane is bubbled through molten metal, with In a fi rst-of-its-kind design, Kawasaki Heavy Industries hydrogen gas and par culate carbon emi ed from the (KHI) in collabora on with Shell, is developing a purpose- top. A newer development by BASF uses a proprietary built liquefi ed hydrogen tanker capable of shipping 3 catalyst in a fl ow-through system in which hydrogen 1250 m (88,500kg) liquid hydrogen (LH2) from Victoria exits the top and carbon falls out of the bo om of the to Japan in 16 days. According to the New Energy reactor.5 Rela vely high carbon effi ciencies are reported and Industrial Technology Development Organiza on for both systems (above 90%), but the energy demand (NEDO), the carrier is forecast to be ready to make its is substan al, with an es mated effi ciency of 55% by fi rst shipment in 2020/2021. These sorts of op ons Wegen et al.4 However, as compared to systema c will signifi cantly increase the costs of gas but may s ll changes in industrial processes, this energy cost could be drama cally less costly than wholesale revision of make sense at industrial scale. Detailed system analysis industrial processes. An important innova on pathway is is required. to make a detailed comparison of the transi on costs to determine if large-scale eff orts at rethinking the use of Hydrogen proponents point out that one of the gas for industrial purposes is warranted in a low-carbon signifi cant drawbacks to natural gas usage—leakage future. from pipelines—could be signifi cantly limited by transforming the natural gas to hydrogen near its Hydrogen requires another set of safety and use source and transmi ng hydrogen in the pipelines with approaches if it is to be used in gas pipelines, including a much smaller impact on climate if leakage occurs. material compa bility, fl ame awareness (hydrogen Combina ons with solar thermal—both as a high- is invisible when burning) and odoriza on. Gasket temperature-process heat source for the methane materials need to be evaluated, and geologic storage of pyrolysis and as a heat storage system for combining hydrogen gas needs to be experimentally demonstrated. other renewable energy—are an a rac ve innova on pathway. Other biofuels This innova on pathway envisions a mix of gases, Biomass provides a commonly used approach to replace renewable natural gas and hydrogen, playing the role fossil fuels. Solid forms like torrefi ed biomass can be that natural gas does today. Other promising technical used directly in place of coal. Bio-char, while commonly approaches include low-cost electrolyzers and nuclear considered as a soil amendment, could be used instead

December 2019 54 as a fuel. These op ons are par cularly interes ng for Of great interest in this area is the possibility of the calciner in cement plants, which are renowned for combining biomass heat sources with CCS, yielding a their ability to burn almost any fuel. Bio-oil is created by por on of nega ve emissions. In this scheme a cement fast pyrolysis of biomass and can also be used as a liquid kiln could rela vely easily operate the calciner on fuel in many systems that use hea ng oil today, although biomass fuel, while capturing and sequestering some

some hydrogena on is required to stabilize the bio-oil. frac on of its total CO2 emissions and thereby off se ng all of its emissions. For many industrial processes, par al These op ons suff er from the same availability issues as capture on exis ng equipment is much more prac cal renewable natural gas. In most integrated assessment than 100% capture. model evalua ons, there is insuffi cient biomass to meet all the compe ng needs in a decarbonized world. A major innova on pathway issue is evalua ng the Innovation Agenda for Zero-Carbon Fuels compara ve costs of transi on for using this resource Innova on pathways for zero-carbon fuels focus in industry. Because these fuels can be used with only on produc on of the fuels, systema c transport minor process changes, they are a rac ve for some and distribu on issues associated with large scale industrial sectors as a means to decarbonize. The replacement of natural gas by renewable gases, and scale is realis c. It takes about 200 kg of coal to create more focused replacement of coal by biomass. one ton of cement. Replacing that coal with torrefi ed ■ Evalua on of the cost benefi ts and life cycle of the biomass—notwithstanding constraints on fuel hea ng replacement strategies value in the kiln—would require about 2-3 mes as ■ Evalua on of transporta on and distribu on methods much original dry biomass. In California, that would and costs require about 4-5 million tons of biomass to fuel the ■ Development of safety and use schemes state’s cement produc on. Recent es mates indicate ■ that about 70 million tons of dried biomass could Improved produc on schemes and assured zero- be obtained from waste sources in California.7 Using carbon technologies torrefi ed biomass to eliminate the energy emissions (but ■ Evalua on of materials for transport and storage of hydrogen, including odorants and fl ame visualiza on. not the calcining emissions) would reduce CO2 emissions from cement produc on by about 50%. Thus, biomass ■ Evalua on of the ming and staging of adding for replacement of coal industrial heat, par cularly in frac onal amounts of zero-carbon fuels to industrial cement produc on, which is rela vely tolerant of fuel systems. quality, is an innova on pathway that can be considered along with the use of biomass to produce liquid fuels. Improved heat application The major innova on needs in this area are to develop Electrification effi cient means of conver ng biomass to transportable Use of electricity to provide heat in industrial processes forms without more carbon emissions—for instance, requires massive changes in industrial equipment as torrefac on is o en done today with external natural discussed in Chapter 2C. Three major issues arise: gas hea ng. Autopyrolysis systems that use the biomass itself to provide heat would avoid those emissions. 1. There has been li le R&D on massive electrifi ca on op ons. A major innova on issue is the extent to which na ons may choose to become biomass exporters to provide 2. Transfer of heat in systems such as blast furnaces is a industrial heat sources. The opera ons at the Drax func on of the physical dimension of the coke, which power sta on in England have demonstrated that a supplies support as well as heat. large-scale, long-distance biomass supply chain can be 3. Electrifi ca on can rarely be done in the context of created and sustained—although the carbon emissions exis ng equipment. from this supply chain may be substan al. As with Dealing with the second issue is par cularly diffi cult. The hydrogen, new infrastructure pathways could make highly tuned nature of the chemical reac ons and heat delivery of biomass as an industrial heat source a global transfer in a system like a blast furnace makes it unlikely commodity. to be decarbonized by either electricity or biomass; the most likely approach is CCS (below) or change to

December 2019 55 completely diff erent processes where the chemical- literature (e.g.,8,9) indicates that this is possible, but reducing poten al of either hydrogen or biomass can be currently will use an enormous amount of electricity. u lized. Innova on around more effi cient electrochemical processes, par cularly reduc on in overvoltage and The third issue in the list above is more amenable to an resistance losses, is needed. innova on pathway, as discussed in the electrifi ca on chapter. The applica on method for electricity to Innovation agenda for improved heat industrial processes can be changed drama cally by application using direct microwave or induc ve hea ng to deposit energy. Each industrial process requires R&D to Innova on pathways for electrifi ca on focus on the iden fy the most effi cient electrical energy deposi on determina on of the most cost-eff ec ve approaches. method, and there is rela vely li le room for generic ■ Evalua on of the most appropriate electric energy development. Each industry must be considered and the deposi on methods by industry. choice of new electrical hea ng method evaluated. This ■ Improving combus on systems, burners and requires very diff erent approaches than for revisions to combustors. the gas grid, since every industrial process will require ■ Evalua on of the compara ve capital costs of slightly diff erent innova on pathways. reconfi guring each industrial process, which will be Improved electrifi ca on pathways have had extremely very substan al and must be weighed against CCS or li le research a en on. Basic understanding of how zero-carbon fuel op ons. Among the key industrial heat is deposited in material, skin depths for dielectric processes which will have signifi cantly specialized hea ng, changes in resistance with chemical changes, electrifi ca on methods are: and safety issues have been evaluated at small scale but 4. Calcining rarely at the size of industrial processes. Na onal-level 5. Refi ning programs addressing key industries are required to fi ll 6. Dis lla on this gap. 7. Glass produc on Solar and stored heat 8. Removing impuri es from metals 9. Ceramics produc on Many research teams are focused on solar thermal 10. Drying applica ons to industrial processes. With focusing mirror systems that can readily exceed 1,000 °C, this ■ Development of more effi cient high-temperature heat is academically a rac ve but requires appropriate storage. heat transfer mechanisms to move the heat from ■ Development of new electrochemical methods that the mirror system to the industrial process. Typically, require much less electricity than today’s incipient industrial processes at this temperature use direct methods. combus on hea ng, so there is no exis ng art around ■ The innova on need for solar and stored heat is the movement of heat at these temperatures. However, around be er heat transfer systems and fl uids for the demonstrated ability to store heat in solar-thermal moving heat at those high temperatures. electricity genera on makes that op on interes ng to ■ The innova on agenda for reduced heat focuses on industry. new electrochemical methods that require much less electricity than today’s incipient methods. Reducing or eliminating heat in processing An obvious effi ciency pathway is to develop catalysts Carbon capture and storage (CCS) and hybrid that replace heat as a means of speeding chemical approaches processes. Par cularly in the chemical industry, heat has been the method of choice for speed. Innova on Carbon capture and storage in catalysis will be valuable there. A major innova on CCS has not been extensively applied to industry. In would be the replacement of thermochemical Chapter 2D we outlined the major opportuni es, which produc on systems with electrochemical methods, are focused on R&D to demonstrated specifi c CCS

par cularly those that begin with CO2. A growing methods for industry. The major challenges are mul ple

December 2019 56 small emi ers in single facili es like refi neries and the and storing CO2 before it reaches the industrial facility. need for specifi c process designs for each industry Bioenergy with CCS (BECCS) is a likely component of type. A substan al demonstra on program is required hybrid systems. to reduce the technology risk for each industry’s CCS Timing issues will be important for hybrid systems. For approach. instance, 10% subs tu on of hydrogen is straigh orward Applica on of CCS to industry will also require the in most burner systems, but higher amounts eventually collec on and geologic storage systems that electric require new burners and new delivery materials. power systems need. This is an opportunity for na onal Similarly, the other uses of a gas grid need to be carbon management programs to consider infrastructure considered—will industry share transi on costs with needs like pipelines, but these needs are not specifi c to home and business gas use, or will those be electrifi ed industrial applica on of CCS. independently? Oxygen-fi red systems are a likely component of The fi nal novel hybrid pathway is to use nega ve decarbonized industrial heat. The produc on and emissions—primarily direct air capture—to off set transporta on of oxygen for these systems is a industrial emissions. At some level it is extremely likely signifi cant challenge, also not unique to this report, but that this will be needed, as it is clear that complete of a similar scale to the produc on of hydrogen in terms decarboniza on of industrial heat is diffi cult. Today of impact and complexity. costs for direct air capture systems are approximately 10 $600/ton CO2. It is widely expected that these costs Hybrid approaches can be reduced to the vicinity of $200/ton, and the Throughout this report we have highlighted the developers believe that $100 can be obtained. For an diffi culty of decarbonizing industrial heat due to the industrial facility opera ng on exis ng equipment with great variety of systems that must be decarbonized 90% capture, it may be cost-eff ec ve to capture the and the limited number of op ons for performing remaining 10% of the emissions at a large air-capture that task. Our discussion has focused on individual facility shared with other facili es. solu ons—hydrogen, electrifi ca on, biofuels, etc.—but hybrid approaches may also be very useful. These Innovation agenda for CCS and hybrid include par al fuel decarboniza on, par al changes approaches and electrifi ca on of the produc on environment, and CCS needs specifi c design studies for applica on par al addi on of CCS to par cularly suscep ble parts of to industrial streams. The opportunity for hybrid the produc on scheme that con nue to use fossil fuels approaches in this area is large, requiring the because of their par cular suitability. These solu ons development of robust cost and life-cycle models to may start as par al decarboniza on that is highly cost es mate the costs of combined approaches, including eff ec ve and evolve into full decarboniza on as mul ple those that use shared facili es. Another innova on approaches are applied. For instance, improved burner pathway is to consider if there are op ons to combine technology may be incorporated as par al hydrogen direct air capture, which is very dependent on heat, or renewable gas become available, or electrifi ca on with industrial waste heat sources.11 While the fi rst of small-scale heat sources may occur inside refi neries priority should always be to reduce waste heat to the while larger units like steam methane reformers depend greatest extent possible, where this waste heat is of a upon CCS. low quality and not economically feasible to recover, it could be used for solid-sorbent-based processes (e.g., in Hybrid approaches may be par cularly useful in combina on with heat pumps). This could be par cularly combining CCS on exis ng equipment, with par al useful to deal with emissions from the “use phase” of biofuel subs tu on on that exis ng equipment. Because products that are not amenable to CCS. The innova on of the carbon neutral nature of the biofuel, doing agenda is to consider the capital and opera ng cost par al carbon capture can off set the emi ed fossil-fuel cross-over point where direct air capture becomes the component, resul ng in net zero emissions. Such more eff ec ve way to achieve addi onal decarboniza on schemes with a biomass component have many fl avors, and whether large-scale systems could be u lized to including gasifi ca on of biomass to hydrogen, capturing polish the emissions of en re industrial sectors.

December 2019 57 Near term Mid term Long term Improve Evaluate cost Research Improve H renewable 2 Zero Carbon benefi ts and power-to-gas Develop safety Determine pipeline materials for natural gas & Fuels transport (H and CH ) methods for H H limits transport 2 4 2 2 H produc on costs technologies 2 and safety methods Evaluate Develop more Improved Evaluate costs Improve exis ng electricity effi cient heat Heat of reconfi guring combus on deposi on storage and Applica on processes systems methods transport

Design CCS Evaluate cost Evaluate costs of Test hybrid Hybrid for specifi c and lifecycle for off -site carbon biomass/capture industries hybrid processes management systems

Develop Evaluate costs Evaluate costs Determine Develop biomass large-scale and impacts and impacts where to Demonstrate large- systems that are Cross- hydrogen of large-scale of large-scale commit exis ng scale hydrogen carbon-nega ve Cu ng produc on hydrogen biomass biomass shipping across the life- and transport transport resources cycle transport Table 4-1. Major Innova on Pathways

Cross-cutting issues: system innovation and Power-to-gas is a second pathway that requires cross- transportation pathways for biomass and cu ng development and agreement. Will we a empt hydrogen to repower industry with a zero-carbon gas system? What methods would be used to create the hydrogen or As we move toward a decarbonized industrial system, renewable natural gas? Today there is no clear economic society will have to make choices in two major areas: pathway to evaluate this op on, mainly because how to use available biomass, and whether to use a en on has been focused on the electric power sector. renewable power to create gas that can be distributed in As renewables become cheaper, use of hydrogen and exis ng networks. renewable natural gas is less a rac ve for electricity As we saw in the biomass chapter, it is possible to genera on but may s ll be of primary importance for imagine powering a substan al por on of tomorrow’s industry. However, a reduced gas grid for industry use industry with biomass, but that use would be in s ll needs to be maintained at a considerable cost. A compe on with liquid fuels, pure nega ve emissions clear understanding of what it would take to create a and, of course, the uses for biomass today, including zero-carbon gas system is necessary to make it possible cooking and hea ng. Most authors today prefer to to consider this op on. consider only the amount of biomass that would be And just as ocean-shipping of liquid natural gas (LNG) available from waste and supplies of biomass that do not rapidly changed the energy world, worldwide shipping place pressures on the availability of food or ecosystem of biomass and hydrogen could have a similar eff ect for services. That amount of biomass is therefore limited industry. Na ons and regions could choose to commit but extremely valuable from a climate perspec ve. resources to large-scale produc on of these heat Today there is no clear metric to decide where to sources and send them by ship to industrialized regions. commit our biomass resources—crea ng such a metric The off se ng eff ects of inves ng in new infrastructure is a major innova on pathway. Interna onal agreement to ship these products and of maintaining expensive will be important, since the Drax experiment in the UK industrial facili es with minimal changes could make has demonstrated that large-scale biomass shipping that an economic- and climate-appropriate solu on for is feasible, even if it does not achieve the desired many forms of industrial heat. The me to evaluate the economic and climate impact.

December 2019 58 possible impacts of such choices is now, before massive 4 Wegen, L. et al. (2017) Methane cracking as a bridge investments are made. technology to the hydrogen economy. Energy Volume 42, Issue 1, 5 January 2017, Pages 720-731 h ps://doi. These innova on pathways are combina ons of applied org/10.1016/j.ijhydene.2016.11.029 and basic research, refl ec ng the deep need for new 5 Houlton, Sarah (2019) Curbing industry carbon processes and approaches in industry. Past roadmaps emissions. BASF methane pyrolysis h ps://www. have called out separate innova on agendas for these chemistryworld.com/news/curbing-industrys-carbon- topics, but the need for broad changes suggests a great emissions/3010737.ar cle ChemistryWorld July 16, number of basic and applied topics. We must move 2019. further down the innova on pathways before there are 6 CSIRO (2019) Australia hydrogen roadmap. h ps:// clearly defi ned needs in specifi c areas. www.csiro.au/en/Do-business/Futures/Reports/ Analysis and modeling are key elements of all these Hydrogen-Roadmap innova on pathways and are more important in the 7 Williams, Robert B, Bryan M. Jenkins Steve Ka a hybrid and new-fuel scenarios such as shipping large (2015) An Assessment of Biomass Resources in amounts of biofuel long distances. Indices of success California, 2013 (DRAFT) Public Interest Energy are needed, and tools to compare op ons across broad Research (PIER) Program INTERIM PROJECT REPORT. h ps://biomass.ucdavis.edu/wp-content/uploads/ swaths of industry must be developed. This need for CA_Biomass_Resource_2013Data_CBC_Task3_DRAFT. compara ve and planning tools is the founda onal pdf element of all the innova on pathways described here, 8 as is the need for broadly applicable data about heat use P De Luna, C Hahn, D Higgins, SA Jaff er, TF Jaramillo, EH Sargent (2019). What would it take for renewably in industry. Without these, innova on cannot proceed in powered electrosynthesis to displace petrochemical the most effi cient and diverse ways. processes? Science 364 (6438). 9 A Kätelhön, R Meys, S Deutz, S Suh, A Bardow (2019) 1 IEA (2019) Hydrogen, tracking clean energy progress. Climate change mi ga on poten al of carbon capture h ps://www.iea.org/tcep/energyintegra on/hydrogen/ and u liza on in the chemical industry Proceedings 2 Bullis, Kevin (2013) Audi to Make Fuel Using Solar of the Na onal Academy of Sciences 116 (23), 11187- Power MIT Technology Review, h ps://www. 11194 technologyreview.com/s/510066/audi-to-make-fuel- 10 Pacala et al. (2018) Na onal Academies of Sciences, using-solar-power/ Engineering, and Medicine. Nega ve Emissions 3 Hafenbradl ,Doris and Mich Hein (2015) Power-to-Gas Technologies and Reliable Sequestra on: A Research – A solu on for energy storage. Gas for Energy, Issue 4 Agenda. Washington, DC: The Na onal Academies 2015. h ps://www.gas-for-energy.com/fi leadmin/G4E/ Press. doi: h ps://org/10.17226/25259. pdf_Datein/g4e_4_15/gfe4_15_ _Hafenbradl.pdf 11 Jennifer Wilcox Personal Communica on (August 2019)

December 2019 59 and others, many governments will be reluctant to impose CHAPTER 5 policies that disadvantage domes c companies in these industries in interna onal trade or might lead companies to shi produc on abroad. This will constrain the set of POLICY poli cally acceptable policy responses in many cases. Policy tools are essen al for decarbonizing industrial Many policy tools are available to help with heat, both in the short- and long-term. This chapter decarbonizing industrial heat. These are discussed discusses the ra onale for policy support and range of below. policy tools available. Policy tools Rationale Government support for R&D The concentra on of carbon dioxide (CO ) in the 2 Na onal governments spend roughly $15 billion annually atmosphere is higher than at any me in human history. on R&D for clean energy technologies. These programs Human ac vi es, including fossil fuel combus on and have played important roles in the development of deforesta on, con nue to increase that concentra on. countless technologies in recent decades.8 The impacts include heat waves, more severe and frequent storms, sea-level rise, forest loss, and ocean Several recent government R&D programs have targeted acidifi ca on.1 innova ons in industrial heat. These include: These problems are classic “externali es.” Market 1. A US Department of Energy (DOE) ARPA-E program 9,10 forces alone will not control CO2 emissions adequately, on novel heat-exchanger technologies. since emi ers of CO2 do not bear the full costs of their 2. A French Na onal Center for Scien fi c Research emissions. Government policies are essen al.2,3 program on high-temperature solar-heated reactors 11 Refl ec ng this, more than 175 countries have ra fi ed for industrial produc on of reac ve par culates. the Paris Agreement, which requires each of them 3. A European Commission program on integra on to regularly report on their policies for controlling of solar heat in industrial processes of the agro- food industry.12 Increased funding for R&D emissions of CO2 and other heat-trapping gases. The Paris Agreement calls for “holding the increase in the on decarbonizing industrial heat could speed global average temperature to well below 2 °C (3.6 °F) deployment and yield important benefi ts. This above pre-industrial levels,” “pursuing eff orts to limit roadmap iden fi es a number of priority areas for the temperature increase to 1.5 °C (2.7 °F) above pre- R&D investment. (See Chapter 4 above.) industrial levels” and achieving net zero emissions in the In December 2015, heads of state from more than 20 4 second half of this century. countries announced Mission Innova on, a coali on Decarbonizing industrial heat is an important part dedicated to accelera ng clean energy innova on. of any strategy for mee ng those goals. As noted in Member governments (including Japan, China, the United Kingdom, Germany and Saudi Arabia) pledged previous chapters, roughly 10% of global CO2 emissions come from produc on of industrial heat—more than to double R&D on clean energy within fi ve years. emissions from cars and planes.5,6,7 Achieving the goals The increase in R&D budgets from these countries set forth in the Paris Agreement would be diffi cult if not in the years ahead off ers an opportunity to increase impossible without cu ng emissions from industrial government R&D funding for decarboniza on of heat produc on. industrial heat, including in the areas above. This will present signifi cant challenges. Several industries The US helped launch Mission Innova on and remains a member. Although the US is unlikely to fulfi ll its overall with high CO2 emissions from heat produc on (including iron, steel and some chemicals) are strategically doubling pledge under the Trump administra on, the US important to host governments and exposed to Congress has increased funding for energy effi ciency and compe on from foreign trade. Some of these industries renewable energy programs at the US DOE in each of the provide considerable employment. For these reasons past several years.13

December 2019 60 Government Procurement for such tax incen ves. They include: ■ In many countries, government procurement makes Investment tax credits: Governments could up more than 10% of GDP.14 Government purchases provide businesses a tax credit for a percentage can play an important role in star ng and building of the capital costs incurred in transi oning to new product markets. First, government purchase low-carbon industrial heat. (This would be similar contracts can provide developers and manufacturers to the US federal government’s investment tax of new products with an assured market, which can be credit for solar power, which has historically especially important in securing debt capital. Second, provided a tax credit of 30% of the cost of any government purchases can help establish standard solar installa on in the US.) technical specifi ca ons for new products, which can help ■ Produc on tax credits: Governments catalyze effi cient supply chains. could provide a tax credit for any products manufactured using low-carbon industrial heat. Governments are major purchasers of steel, cement, (This would be somewhat similar the US federal chemicals and other products that require heat in government’s produc on credit for wind power, the manufacturing process. Procurement standards which provides a tax credit based on the kWh of that give preferences to products with the lowest wind power sold at a facility.) embedded carbon content could drive signifi cant ■ Waiver of sales, value-added taxes or import changes in industrial behavior. Procurement standards taxes: Governments could waive taxes that that authorize purchasing offi cials to base decisions would otherwise be imposed on any products on lifecycle carbon emissions of products could do the manufactured using low-carbon heat. (This would same. California’s Buy Clean statute is a leading example be similar to Norway’s incen ves for electric of legisla on that directs authori es to pay a en on to vehicles, which include waivers of import and climate impacts in the procurement process. Similarly, sales taxes that apply to conven onal vehicles.) procurement regula ons could give preferences to B. Grants. Grants are among the most direct ways products manufactured without the use of fossil fuels to to provide fi nancial support for the low-carbon generate heat. transi on. Grant programs are widespread in many Fiscal subsidies countries, o en to assist in deployment of fi rst-of- a-kind or early-stage technologies. Governments Decarbonizing industrial heat will impose costs on could provide grants to help defray the capital costs aff ected businesses. Capital expenditures may be associated with the transi on to decarbonizing required to retrofi t facili es or build new facili es. The industrial heat processes. cost of physical assets may need to be wri en off if C. Loan guarantees. Cu ng the cost of debt capital can those assets are re red before the end of their useful help make a project fi nancially viable. Government lives. Opera ng expenses may increase if inputs are loan-guarantee programs seek to do that by reducing more expensive than current fossil fuel inputs. risk to lenders, resul ng in lower borrowing costs. Government policies can help to reduce those costs with The US DOE’s loan-guarantee programs helped fi scal subsidies. These can take several forms. Leading launch the u lity-scale solar industry in the US, op ons are discussed below. among other successes. Loan guarantees for the capital expenditures required for decarbonizing A. Tax Incen ves. Tax incen ves can play an important industrial heat could signifi cantly speed deployment. role in spurring deployment of clean energy products. In Norway, for example, generous tax D. Feed-in tariff for renewable natural gas. A incen ves helped plug-in electric vehicles capture feed-in-tariff provides a guaranteed price for an 50% of new car sales in 2018.15 In the US, federal energy product for a set number of years. This can tax incen ves have played an important role in drama cally improve bankability of projects and help promo ng deployment of solar and wind power. to scale up produc on. Feed-in-tariff s have been Such incen ves could play a similar role in promo ng used to help launch markets for solar power and alterna ves to the use of fossil fuels in industrial other renewables around the world. Germany and heat produc on. There are many possible structures the Netherlands have implemented feed-in tariff s for

December 2019 61 biogas. This is an important tool for helping promote among academic economists as the most cost-eff ec ve produc on and use of renewable natural gas. approach for addressing climate change.17 E. Contracts for Diff erences. Contracts for Diff erences Carbon pricing con nues to grow steadily around the are used in the United Kingdom to support low- world. Fi y-seven jurisdic ons covering 20% of global carbon electricity genera on. In a Contract for greenhouse gas (GHG) emissions now have carbon Diff erences, the government guarantees a power pricing ini a ves that have been implemented or are supplier will receive a stated amount, covering the scheduled for implementa on.18,19 diff erence between that amount and the amount the power supplier actually receives. This policy could be Unfortunately the results of those programs are used to help support deployment of low-carbon heat not encouraging in several respects. First, very few technologies as well.16 carbon-pricing programs have resulted in carbon prices suffi cient to signifi cantly reduce emissions. Governments Low-carbon product standards have generally been unwilling to impose such prices, A low-carbon product standard sets a limit on o en due to strong opposi on from the businesses and the product’s life cycle emissions. Low-carbon individuals most exposed to energy price increases. fuel standards—the leading example of such an Second, programs in several jurisdic ons (including approach—have been adopted in California, Oregon, those in Australia and several US states) have been Bri sh Columbia and the European Union. California’s suspended or terminated due to changes in poli cal low-carbon fuel standard requires producers of leadership. petroleum-based fuels to reduce the carbon intensity of The carbon prices that might be needed to induce a their fuels 10% from 2010 levels by 2020. The EU’s Fuel transi on from fossil fuels for industrial heat produc on Quality Direc ve requires reduc ons of 6% in the carbon are unclear. These levels likely vary considerably from intensity of fuels from 2010 levels by 2020. industry to industry and even facility to facility. In the Such standards could be applied to a range of products absence of readily available subs tute processes for currently manufactured using fossil fuels to generate genera ng heat, the carbon prices required to induce heat. The administra ve complexi es associated with a transi on from fossil fuels would likely be quite high. such a program could be considerable, in part because Few if any countries have demonstrated a willingness to many of the most relevant products are inputs into other set carbon prices at these levels. products. However low-carbon product standards could Nevertheless, a growing number of businesses use provide considerable incen ve for manufacturers to fi nd “shadow carbon prices” when making long-term capital alterna ve ways of genera ng such heat. investment decisions. (That is, the businesses apply a carbon price in calcula ng returns on capital, even if Infrastructure development no carbon price or a lower carbon price is imposed in The transi on to low-carbon industrial heat may require the jurisdic on in which they operate.) These shadow new infrastructure (such as electric transmission lines carbon prices may play an important role in aff ec ng or hydrogen pipelines). Governments can play a central capital investment decisions on the margin. The role facilita ng the development of such infrastructure adop on of mandatory carbon pricing programs in through permi ng, fi nancing and other measures. jurisdic ons around the world, even with carbon prices Governments can also take a direct role through at modest levels, may encourage business use of shadow development and ownership of such infrastructure carbon prices. where it serves a common good—as is o en the case for road, rail, district hea ng and water infrastructure. Carbon tariffs In 2017, 440 million tons of steel was traded Carbon prices interna onally. This was more than a quarter of global A price on carbon dioxide emissions, whether through produc on.20 Other products that require heat in the an emissions trading program or tax mechanism, manufacturing process—including some chemicals—are provides emi ers with an important incen ve to cut traded interna onally in high volumes as well. Many emissions. Carbon pricing enjoys overwhelming support governments may be reluctant to impose costs related

December 2019 62 to decarbonizing industrial heat on the manufacture of Mandates such products, due to concerns about disadvantaging Governments mandates can be eff ec ve in helping build such products in interna onal trade. markets for clean energy products. In the US, many state Carbon tariff s (some mes called “carbon border tax governments require u li es to purchase a minimum adjustments”) are a tool for addressing that concern. A percentage of their power from renewable sources. In country that requires its manufacturers to transi on to India, a similar requirement is imposed by the Ministry low-carbon industrial heat could tax imports of relevant of New and Renewable Energy. These requirements have products from countries that fail to do so. This could been important to the early growth of wind and solar level the playing fi eld, elimina ng the disadvantage power in both countries.25 domes c manufacturers face from higher costs Other experiences suggest cau on, however. The US associated with decarbonizing their hea ng processes. federal government mandate has required the use of No carbon tariff s have ever been adopted. There at least cellulosic ethanol in fuel supplies for almost a decade. three prac cal concerns with carbon tariff s: Nevertheless, the cellulosic ethanol industry remains in its infancy and waivers to that requirement have ■ Such tariff s may not be legal under the rules of the been granted on a regular basis. Technology-forcing World Trade Organiza on (WTO). As a general rule, requirements—in which governments require private the WTO prohibits restric ons on the import of goods actors to meet standards that are not yet technically based on anything other than a ributes of goods at achievable—have been successful in some instances but the border. (Countries are not allowed to discriminate not in others.26 between goods based on characteris cs of upstream manufacturing processes.) There are excep ons, Government mandates could help spur the transi on to including some that might apply to carbon tariff s (such low-carbon industrial heat. Governments could prohibit as an excep on related to environmental protec on). the use of fossil fuels in genera ng heat in certain In recent years the permissibility of carbon tariff s industrial sectors a er a certain date, for example. under the WTO has been debated extensively by Or governments could require the use of low-carbon leading trade experts. The issue has not been resolved industrial heat technologies a er a certain date. by a WTO tribunal.21-24 Business investment in compliance strategies with such ■ Challenging design and administra ve ques ons must mandates could help to spur innova on. be addressed to implement a carbon tariff program. Decisions must be made about which products are Voluntary industry associations subject to the tariff s. (For example: Would steel from a Industry associa ons such as World Steel Associa on, zero-emissions steel plant located in a country without World Petroleum Council, World Cement Associa on and carbon emissions limits be subject to the tariff ?) World Business Council on Sustainable Development can Decisions must also be made about how to set carbon help develop methods and standards for decarbonizing tariff levels. Depending on those decisions, extensive industries. They can play an important role in data collec on and processing could be required to informa on-sharing on such topics as well. Governments eff ec vely administer the tariff s. can encourage such ac vi es by hos ng mee ngs, ■ Although carbon tariff s could in theory level the providing recogni on and off ering fi nancial support. playing fi eld for manufacturers in their home markets, they do nothing to help manufacturers selling abroad. Clean Energy Ministerial A manufacturer that incurred addi onal costs to The Clean Energy Ministerial is a global forum where decarbonize industrial heat processes would s ll be major economies work together to share best prac ces at a cost disadvantage in foreign markets. Other tools, and promote policies and programs that encourage such as cost rebates for exports, would be required to and facilitate the transi on to a global clean energy ensure level playing fi elds abroad. economy. A Clean Energy Ministerial ini a ve on industrial heat decarboniza on could help to share best prac ces and accelerate their adop on. Any country that

December 2019 63 par cipates in the Clean Energy Ministerial could launch 8 Mission Innova on, “Tacking Progress,” Site accessed such an ini a ve. on 11/27/19, h p://mission-innova on.net/our-work/ tracking-progress/. Decarboniza on of industrial heat is unlikely to happen 9 at scale as a result of voluntary measures. Policy ARPA-E, “DOE Announces $36 Million for High- Temperature Materials Projects,” ARPA-E News supports will be required for this transi on. The diversity (2019), h ps://arpa-e.energy.gov/?q=news-item/ of industries in which supports are needed complicates doe-announces-36-million-high-temperature-materials- policy design and implementa on. More research, projects analysis and consulta on with key stakeholders are 10 ARPA-E, “ARPA-E Announces Funding Opportunity to needed to shape the best policies to help meet this Develop Novel Heat Exchanger Technologies,” ARPA-E challenge. News (2018)h ps://arpa-e.energy.gov/?q=news-item/ arpa-e-announces-funding-opportunity-develop-novel- 1 IPCC, 2018: Summary for Policymakers. In: Global heat-exchanger-technologies Warming of 1.5°C. An IPCC Special Report on the 11 NKS Energie, “Funded Projects under Horizon impacts of global warming of 1.5°C above pre-industrial 2020: Secure, clean and effi cient energy,” levels and related global greenhouse gas emission European Union Open Data Portal (2016) pathways, in the context of strengthening the global at p. 13, h ps://www.nks-energie.de/ response to the threat of climate change, sustainable lw_resource/datapool/systemfi les/elements/ development, and eff orts to eradicate poverty fi les/5B5C11BA3DC812B5E0539A695E86ECA3/current/ [Masson-Delmo e, V., P. Zhai, H.-O. Pörtner, D. Roberts, document/H020_-_Low_Carbon_Energy_Projects_ J. Skea, P.R. Shukla, A. Pirani, W. Moufouma-Okia, funded_2014.pdf C. Péan, R. Pidcock, S. Connors, J.B.R. Ma hews, Y. Chen, X. Zhou, M.I. Gomis, E. Lonnoy, T. Maycock, M. 12 Ship2Fair, Site accessed 11/27/19, h p:// Tignor, and T. Waterfi eld (eds.)]. World Meteorological ship2fair-h2020.eu/. Organiza on, Geneva, Switzerland, 32 pp., h ps:// 13 Adrian Cho, “Congress sends Trump energy spending www.ipcc.ch/sr15/ bill that includes healthy boost for science,” Science, 2 Sir Nicholas Stern, “Stern Review on the Economics of (2018), h ps://www.sciencemag.org/news/2018/09/ Climate Change,” The Na onal Archives: HM Treasury, congress-sends-trump-energy-spending-bill-includes- (2006), h ps://webarchive.na onalarchives.gov.uk/+/ healthy-boost-science h p://www.hm-treasury.gov.uk/sternreview_index.htm 14 Esteban Or z-Ospina and Max Roser, “Public Spending,” 3 Alison Benjamin, “Stern: Climate change a ‘market Our World in Data (2017), h ps://ourworldindata.org/ failure,’” Guardian (2007) (quo ng Sir Nicholas Stern public-spending/ –”Climate change is a result of the greatest market 15 Akshat Rathi, “Half of all cars sold in Norway in 2018 failure the world has seen”), h ps://www.theguardian. were electric,” Quartz (2019), h ps://qz.com/1514111/ com/environment/2007/nov/29/climatechange. half-of-all-cars-sold-in-norway-in-2018-were-electric/ carbonemissions 16 Department for Business, Energy & Industrial Strategy, 4 Paris Agreement Ar cles 2(1)(a) and 4(1) “Contracts for Diff erence: Policy Paper,” Gov.UK 5 IEA, “CO Emissions Sta s cs: An essen al tool for (2019), h ps://www.gov.uk/government/publica ons/ analysts and policy makers,” Site accessed 11/27/19, contracts-for-diff erence/contract-for-diff erence h ps://www.iea.org/sta s cs/co2emissions/. 17 George Akerlof, et al., “Economists’ Statement on 6 Colin McMillan et al., “Genera on and Use of Thermal Carbon Dividends,” Climate Leadership Council (2019), Energy in the U.S. Industrial Sector and Opportuni es h ps://www.clcouncil.org/economists-statement/ to Reduce its Carbon Emissions,” Na onal Renewable 18 The World Bank, “57 Carbon Pricing Ini a ves Energy Laboratory (2016) at p. viii (industrial heat Now in Place Globally, Latest World Bank Report produc on roughly 10% of total US CO emissions), Finds: Press Release,” The World Bank Group h ps://www.nrel.gov/docs/fy17os /66763.pdf (2019) h ps://www.worldbank.org/en/news/ 7 Interna onal Organiza on of Motor Vehicle press-release/2019/06/07/57-carbon-pricing-ini a ves- Manufacturers (OICA), “Climate Change and CO,” OICA now-in-place-globally-latest-world-bank-report-fi nds. (2008), h p://oica.net/wp-content/uploads/climate- print change-and-co2-brochure.pdf at p.6 (cars, trucks and 19 Interna onal Carbon Ac on Partnership (ICAP), buses combined are 16%). “Emissions Trading Worldwide: Status Report,” ICAP

December 2019 64 (2019), Eds. Marissa San karn, et al., Berlin: ICAP, Research Network (SSRN) (2012), h ps://papers.ssrn. h ps://icapcarbonac on.com/en/?op on=com_ com/sol3/papers.cfm?abstract_id=2026879 a ach&task=download&id=613 24 Adele C. Morris, “Policy Brief: Making Border Carbon 20 US Department of Commerce, Interna onal Trade Adjustments Work in Law and Prac ce,” Tax Policy Administra on, Steel Import Monitoring and Analysis, Center: Urban Ins tute & Brookings Ins tu on (2018), “Global Steel Report,” Global Steel Trade Monitor h ps://www.taxpolicycenter.org/sites/default/fi les/ (2018), h ps://www.trade.gov/steel/pdfs/global- publica on/155511/policy_brief_making_border_ monitor-report-2017.pdf at pp. 3 and 11 carbon_adjustments_work_in_law_and_prac ce.pdf 21 Carbon Tax Center, “Border Adjustments,” (Date 25 Tom Kenning, “Revision to ‘single most important Accessed: September 30, 2019), h ps://www. policy’ to drive solar ready for approval,” PV carbontax.org/nuts-and-bolts/border-adjustments/ Tech (2015), h ps://www.pv-tech.org/news/ intersolar-india-revision-to-single-most-important- 22 Jennifer Hillman, “Changing Climate for Carbon policy-to-drive-solar-re Taxes: Who’s Afraid of the WTO?,” German Marshall Fund of the United States: Climate & Energy Paper 26 David Gerard and Lester Lave, “Implemen ng Series 2013 (2013), h ps://www.scribd.com/ Technology-Forcing Regula ons,” Technological document/155956625/Changing-Climate-for-Carbon- Forecas ng and Social Change (2005), h p:// Taxes-Who-s-Afraid-of-the-WTO faculty.lawrence.edu/gerardd/wp-content/uploads/ sites/9/2014/02/18-TFSC-Gerard-Lave.pdf 23 Joost Pauwelyn, “Carbon Leakage Measures and Border Tax Adjustments Under WTO Law,” Social Science

December 2019 65 integrated, making it diffi cult to pursue simple fuel CHAPTER 6 subs tu on without a larger system redesign. Biofuel and hydrogen combus on may be the most promising op ons for the highest temperature applica ons. FINDINGS AND (Although electrical hea ng pathways can generate high temperatures, they would require very large RECOMMENDATIONS capital investments in many industries.) CCUS applied to hydrogen produc on or combus on facili es remains an Decarbonizing industrial heat produc on will require op on and has the benefi t of also managing byproduct innova ng in mul ple sectors. Progress will require process emissions, but sector-specifi c analyses to date a set of ac ons grounded in improved knowledge, are limited. More broadly, there is a lack of analysis on strong analy cal founda ons and support from key the costs, benefi ts and tradeoff s between alterna ve stakeholders. In this chapter, we summarize our key op ons. fi ndings and recommenda ons. Finding 4: Existing options face challenges based Findings on price, performance and viability. Fossil fuels provide the overwhelming majority of industrial heat Finding 1: Emissions from industrial heat production today. Preliminary analysis suggests that all possible limit progress on climate goals. Roughly 22% of alterna ves carry substan ally higher costs, typically greenhouse gas (GHG) emissions come from industry, 50-500% more, and may have even higher system costs and roughly 40% of those emissions are the result of (e.g., due to addi onal infrastructure requirements). burning fuel to generate heat. This places heat-related In some cases, the carbon reduc ons associated with industrial emissions close to 10% of total global GHG an alterna ve are unclear (i.e., grid-based electricity or emissions—more than cars and planes combined. biofuels). Ques ons remain about the ability of some Deep decarboniza on will be diffi cult or impossible op ons to scale. Many of those op ons do not provide without progress in decarbonizing industrial heat suffi ciently high temperatures for some applica ons. sources. In many cases, it is unclear if a par cular approach represents a viable alterna ve at all, as it is unclear Finding 2: The operational requirements and how it might be used to deposit heat where needed in commercial realities of many industries limit specifi c applica ons. opportunities for decarbonization. The industrial sector is actually many diff erent sectors with dis nct products, Finding 5: There appear to be many pathways markets, technologies and opera onal requirements. to improving cost, performance and viability of Many industrial processes require temperatures above low-carbon industrial heat options. Despite the 300 °C and some above 1,000 °C. The narrow margins challenges alterna ves face today, most approaches of the business and high-capital expense of industrial could be drama cally improved. Although the precise facili es frequently require high-capacity factors for magnitude of poten al cost, performance and life-cycle profi tability. Many industrial products are globally traded improvements are unclear, substan al improvements commodi es that are extremely sensi ve to price. Many in system engineering, performance, process are strategic industries whose economic viability is of intensifi ca on, heat recovery, capital cost and capacity considerable importance to host na ons. are possible in most systems. Novel approaches to some industrial processes appear to be able to provide large Finding 3: There are few options today for low- improvements in cost and life-cycle decarboniza on, carbon heat generation for industry. High-temperature although most require further explora on and tes ng requirements and high-capacity factors limit the op ons before scale-up for subs tu on of fossil fuel heat with low-carbon alterna ves. For new facili es and especially for Finding 6: Many potential policy options exist that exis ng facili es, commercially available op ons today could improve the speed and magnitude of industrial face enormous challenges based on both cost and decarbonization and deployment of low-carbon performance. Many industrial processes are highly alternative heat systems. Government procurement may be an especially potent policy tool. Governments

December 2019 66 are major purchasers of cement, iron and steel, and programs. These programs should be created within other industrial products. Procurement rules that give appropriate ministries and should be commensurate priority to products produced in low-carbon processes in scale to R&D programs in electric power and could spur innova on and deployment. Investments in transporta on decarboniza on. Basic, use-inspired R&D through tax incen ves or grants could also have and applied research should receive support, as should a signifi cant payoff . In contrast, economy-wide carbon pilot tests and commercial demonstra ons (ideally in taxes may have limited impact on GHG emissions from partnership with industry and at opera ng facili es). industrial heat produc on. In part, this is due to the Industrial heat decarboniza on should be added as exposure of many industries to compe on from global priori es to Mission Innova on and the Clean Energy trade. All op ons would benefi t from addi onal analysis. Ministerial. ■ Recommendation 3: Governments should iden fy Recommendations and implement a set of policy ac ons to accelerate ■ Recommendation 1: Key stakeholders should decarboniza on of industrial heat, star ng with priori ze industrial heat produc on as a key element “buy clean” procurement standards. Such standards of any climate mi ga on strategy. Governments, are among the most promising and immediately companies and researchers should priori ze ac onable policy tools. Because mul ple approaches characteriza on and analysis of their industrial sectors. will be necessary to successfully deliver deep and Core data and informa on, such as capacity factors, rapid decarboniza on of industrial heat produc on, fuel purchased and facility-based effi ciency, should be governments should assess which policy op ons best gathered and made publicly available. suit their economic, poli cal and natural resource ■ Recommendation 2: Industry-specifi c analy cal base. frameworks and innova on agendas are essen al. Final thoughts Governments and companies together should This Roadmap is an ini al foray into an important and develop new ini a ves and R&D programs to focus on complex topic. A core fi nding of this Roadmap is that industrial-sector decarboniza on with a focus on heat more work is needed on this topic. The urgency of supplies. Governments, academic researchers and climate change requires rapid ac on. More data, input industrial leaders should cooperate to develop new and technology op ons for decarbonizing industrial heat publicly available data, analy cal tools and training are urgently needed.

DISCLAIMER: Roger Aines and Joshuah Stolaroff contributed to the technical evalua ons in this document. The policy recommenda ons were prepared by the other contributors. The technical evalua ons by Dr. Aines and Dr. Stolaroff were prepared as an account of work sponsored by an agency of the United States government. Neither the United States government nor Lawrence Livermore Na onal Security, LLC, nor any of their employees makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any informa on, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specifi c commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily cons tute or imply its endorsement, recommenda on, or favoring by the United States government or Lawrence Livermore Na onal Security, LLC. The views and opinions of authors expressed herein do not necessarily state or refl ect those of the United States government or Lawrence Livermore Na onal Security, LLC, and shall not be used for adver sing or product endorsement purposes. Prepared by LLNL under Contract DE-AC52-07NA27344.

December 2019 67